Advertisement

  • Loading metrics

Open Access

Peer-reviewed

Research Article

FunOrder: A robust and semi-automated method for the identification of essential biosynthetic genes through computational molecular co-evolution

Abstract

Secondary metabolites (SMs) are a vast group of compounds with different structures and properties that have been utilized as drugs, food additives, dyes, and as monomers for novel plastics. In many cases, the biosynthesis of SMs is catalysed by enzymes whose corresponding genes are co-localized in the genome in biosynthetic gene clusters (BGCs). Notably, BGCs may contain so-called gap genes, that are not involved in the biosynthesis of the SM. Current genome mining tools can identify BGCs, but they have problems with distinguishing essential genes from gap genes. This can and must be done by expensive, laborious, and time-consuming comparative genomic approaches or transcriptome analyses. In this study, we developed a method that allows semi-automated identification of essential genes in a BGC based on co-evolution analysis. To this end, the protein sequences of a BGC are blasted against a suitable proteome database. For each protein, a phylogenetic tree is created. The trees are compared by treeKO to detect co-evolution. The results of this comparison are visualized in different output formats, which are compared visually. Our results suggest that co-evolution is commonly occurring within BGCs, albeit not all, and that especially those genes that encode for enzymes of the biosynthetic pathway are co-evolutionary linked and can be identified with FunOrder. In light of the growing number of genomic data available, this will contribute to the studies of BGCs in native hosts and facilitate heterologous expression in other organisms with the aim of the discovery of novel SMs.

Author summary

The discovery and description of novel fungal secondary metabolites promises novel antibiotics, pharmaceuticals, and other useful compounds. A way to identify novel secondary metabolites is to express the corresponding genes in a suitable expression host. Consequently, a detailed knowledge or an accurate prediction of these genes is necessary. In fungi, the genes are co-localized in so-called biosynthetic gene clusters. Notably, the clusters may also contain genes that are not necessary for the biosynthesis of the secondary metabolites, so-called gap genes. We developed a method to detect co-evolved genes within the clusters and demonstrated that essential genes are co-evolving and can thus be differentiated from the gap genes. This adds an additional layer of information, which can support researchers with their decisions on which genes to study and express for the discovery of novel secondary metabolites.

Citation: Vignolle GA, Schaffer D, Zehetner L, Mach RL, Mach-Aigner AR, Derntl C (2021) FunOrder: A robust and semi-automated method for the identification of essential biosynthetic genes through computational molecular co-evolution. PLoS Comput Biol 17(9): e1009372. https://doi.org/10.1371/journal.pcbi.1009372

Editor: Samuel V. Scarpino, Northeastern University, UNITED STATES

Received: April 13, 2021; Accepted: August 23, 2021; Published: September 27, 2021

Copyright: © 2021 Vignolle et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: The FunOrder tool, the relevant database, and the sequences and the FunOrder output of the negative control GCs and the positive control BGCs are available in the GitHub repository (https://github.com/gvignolle/FunOrder). We have also used Zenodo to assign a DOI to the repository: 10.5281/zenodo.5118984.

Funding: This study was supported by the Austrian Science Fund (FWF, https://www.fwf.ac.at/) [P 29556 to RM, P 34036 to CD] and TU Wien (https://www.tuwien.at/) [PhD program TU Wien bioactive]. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Competing interests: The authors have declared that no competing interests exist.

This is a PLOS Computational Biology Methods paper.

Introduction

Secondary metabolites (SMs) are a diverse group of compounds with a plethora of different chemical structures and properties which are found in all domains of life, but are predominantly studied in bacteria, fungi, and plants [1]. SMs are not necessary for the basic survival and growth of an organism but can be beneficial under certain conditions. For example, pigments help to sustain radiation, antibiotics help in competitive situations, and toxins can serve as defensive compounds or as virulence factors [2,3]. Notably, many SMs are used by humankind as drugs and pharmaceuticals, pigments and dyes, sweeteners and flavours, and most recently also as precursors for the synthesis of plastics [4]. The study of the secondary metabolism holds the promise for novel antibiotics, pharmaceuticals and other useful compounds [5].

A major hinderance in the discovery of yet undescribed SMs is the fact that most SMs are not produced under standard laboratory conditions, as they do not serve a purpose for the organisms then. Currently, different strategies are followed to circumvent this problem [6,7]. Untargeted approaches aim to induce the expression of any SM. To this end, biotic and abiotic stresses are applied, or global regulators and regulatory mechanisms are manipulated [8]. These strategies may lead to the discovery of novel compounds, whose corresponding genes have to be identified subsequently by time-consuming and expensive methods [7]. An extreme example are the aflatoxins, major food contaminants with serious toxicological effects [9]. It took over 40 years from the discovery of the aflatoxins as the causal agent of “turkey X” disease in the 1950s [10] until the corresponding genes were finally described in 1995 [11]. Targeted SM discovery approaches aim to induce the production of specific SMs by either overexpressing genes in the native host or by heterologous expression in another organism [12]. The targeted approaches, also called reverse strategy or bottom-up strategy allows a direct connection of SMs to the corresponding genes and does not rely on the inducibility of SM production in the native host. Inherently, the bottom-up approach is depending on modern genomics and accurate gene prediction tools [13].

In bacteria and fungi, the genes responsible for the biosynthesis of a certain SM are often co-localized in the genome, forming so called biosynthetic gene clusters (BGCs) [14,15]. The BGCs consists of one or more core genes, several tailoring enzymes, and genes involved in regulation and transport. As all these genes are essential for the production of a SM in the native host, we will refer to them as “essential genes” in this study. The core genes are responsible for assembling the basic chemical scaffold, which is further modified by the tailoring enzymes yielding the final SM [16]. We refer to the core genes and the tailoring genes as “biosynthetic genes” in this study. Depending on the class of the produced SM, the core genes differ. In fungi, the main SM classes are polyketides (e.g. the cholesterol-lowering drug lovastatin [17] and the mycotoxin aflatoxin [9]) and non-ribosomal peptides (e.g. the immunosuppressant cyclosporine [18] and the antibiotic penicillin [19]), with polyketide synthases (PKS) or non-ribosomal peptide synthetases (NRPS) as core enzymes, respectively. Other SM classes are terpenoids, alkaloids, melanins [20,21], and ribosomally synthesized and posttranslationally modified peptides (RiPPs) [22,23], whose corresponding genes may also be organized in BGCs. As mentioned, BGCs may also contain genes encoding for transporters [24], transcription factors [25], or resistance genes [26]. While their gene products are not directly involved in the biosynthesis of a SM they are still essential for the biosynthesis; we will call them „further essential genes”in the following and differentiate them from the „biosynthetic genes“. The biosynthetic genes and the further essential genes are both necessary for the biosynthesis of a SM in the native organisms. In contrast, only the biosynthetic genes and a selection of the further essential genes (e.g. transporters) are necessary for heterologous expression [reviewed in [27]]. Notably, fungal BGCs often also contain genes that are not necessary for the production of a SM, the so-called gap genes. The gap genes are not involved in the biosynthesis, regulation, or transport of the SM, but have an unrelated function (Fig 1). We would like to stress here, that this cannot be predicted based only on the class of the gene product. For instance, a gene encoding for a transporter in the aflatoxin BGC was reported to have no significant role in aflatoxin secretion [28].

thumbnail
Download:
Fig 1. Schematic representation of the lovastatin BGC from Aspergillus terreus (lov).

In red the biosynthetic genes for SM production, in gold the further essential genes, and in blue the genes not involved in the biosynthetic pathway.

https://doi.org/10.1371/journal.pcbi.1009372.g001

As mentioned, the bottom-up approach for SM discovery is depending on modern genomics and the accurate prediction of genes and BGCs. Each important gene missing in the prediction is detrimental for obvious reasons, whereas each unnecessarily considered gap gene makes the study of a BGC more complicated and complex, and the construction and transformation processes for heterologous expression more challenging. Currently, several BGC prediction tools are available for fungi. Some tools for genome mining are antiSMASH [29], CASSIS and SMIPS [30], SMURF [31], TOUCAN, a supervised learning framework capable of predicting BGCs on amino acid sequences [32], and DeepBGC, an unrestricted machine learning approach using deep neural networks [33]. These tools are effective and successful in finding and predicting BGCs based solely on genomic data. AntiSMASH uses a rule-based approach to identify BGCs based on the identification of core or signature enzymes and applies a greedy approach to extend a cluster on either side. This may result in overlaps or combinations of closely situated clusters. However, the genes within the predicted BGCs are classified into core biosynthetic genes, additional biosynthetic genes, transport-related genes, regulatory genes, and other genes based on profile hidden Markov models by the antiSMASH tool. The BGC prediction method of CASSIS and SMIPS is based on the principle that the promoter regions of genes in a BGC contain one or more shared motif, as they are co-expressed and presumably regulated by the same regulatory factors and/or mechanisms [30].

As mentioned above, the class of an enzyme may be a good indication for a potential involvement in the biosynthesis of a SM but does not guarantee a correct prediction. This problem can be solved by the analysis of transcriptome data because the genes necessary for SM production within a BGC are normally co-expressed with each other but not with the gap genes [34]. Notably, this demands the knowledge of expression conditions and does not work for silent BGCs. However, it is an obvious advantage to have as much information as possible about a BGC before studying it in the native host or performing heterologous expression for a bottom-up approach for SM discovery.

We speculate that a comparative genomics analysis focusing on the evolutionary history of the genes in a BGC might be a feasible alternative to a transcriptomics analysis in fungi for the following reasons. In general, BGCs are suggested to undergo a distinct and faster evolution than the rest of the genome, based on different mechanisms and genetic drivers [16,35–40]. In bacteria, the evolution of BGCs is strongly influenced by the strong occurrence of horizontal gene transfer in these group of microorganism [39]. Medema et al. performed a large-scale computational analysis of bacterial BGCs and found that many BGCs consist of sub-clusters. These sub-clusters encode for enzymes that work together to form a distinct chemical structure. Notably, this sub-clusters were described as “independent evolutionary entities” and the contained genes are co-evolving. The authors suggested a “bricks and mortar” model. Therein, different sub-clusters, the “bricks” form different chemical building blocks for a secondary metabolite. Additional genes within the BGCs are encoding for enzymes that combine the building blocks, and fulfil other functions such as tailoring, regulation and transport. These individual genes are the “mortar” in the “brick and mortar” model [40]. The “bricks” correspond to what we term “biosynthetic genes” and the “mortar” to our “further essential genes”. Through horizontal gene transfer, the “bricks” can be easily exchanged and recombined to form novel BGCs and secondary metabolites[40]. Notably, not all bacterial BGCs are composed of exchangeable sub-units but some BGCs keep a stable architecture over a long time [40].

In fungi, three molecular evolutionary processes were suggested to be responsible for shaping the BGCs in a recent study, i.e., functional divergence, horizontal gene transfer, and de novo assembly [41]. Rokas et al. define functional divergence as a “process by which homologous BGCs, through the accumulation of genetic changes, gradually diverge in their functions changes” [41] and horizontal gene transfer as a “process by which an entire BGC from the genome of one organism is transferred and stably integrated into the genome of another through non-reproduction related mechanisms” [41]. This implies in both cases, that fungal BGCs are staying intact. Further, the genes are suggested to undergo a co-evolution which is faster than the rest of the genome [41]. Medema’s “brick and mortar” model would more or less correspond to what Rokas et al. describe as “de novo assembly”. This is defined as a “process by which an entire BGC is evolutionarily assembled through the recruitment and relocation of native genes, duplicates of native genes, and horizontally acquired genes” [41]. Notably, Rokas et al. state that this is the”least well-documented evolutionary process involved in the generation of fungal chemodiversity” [41], suggesting that in known and described fungal BGCs functional divergence and horizontal gene transfer are the two main evolutionary process, during which BGCs are staying intact and genes undergo a similar evolution. Further, we hypothesize that especially the biosynthetic genes in a BGC are co-evolutionary linked by the selection pressure to keep the biosynthetic pathway intact. Notably, a co-evolution analysis is a laborious and time-consuming task because a phylogenetic tree has to be calculated for each gene and then the trees compared to each other manually [42]. Recently, a method for the detection of co-evolution in bacterial BGCs was developed with the aim to identify sub-clusters [43]. That method is based on the detection of orthologous genes that are present in close vicinity in many BGCs. This method is working unsupervised but requires a large set of BGCs as input [43].

In this study we describe a method (FunOrder) that allows a fast, semi-automated co-evolution analysis using individual BGCs as input. Based on this analysis and the assumption that the essential genes undergo a shared or similar evolution, FunOrder aims to identify essential genes in BGCs. To this end, we constructed a database of fungal proteomes as basis for the identification of co-evolutionary linked genes in ascomycetes. We determine the thresholds for the detection of co-evolution within different control gene sets. Then, we evaluated FunOrder and tested the underlying hypothesis, whether essential genes within a BGC could be identified based on the principle of co-evolution. We demonstrated the robustness and the applicability of the FunOrder method by analysing different control gene sets, including empirically validated BGCs and evaluated our method using stringent statistical tests.

Material and methods

Construction of a fungal proteome database

In this study we aim to identify co-evolutionary linked genes in ascomycetes. As the basis for the detection of co-evolution is a suitable database [42], we compiled an empirically optimized database consisting of 134 fungal proteomes from mainly ascomycetes and from two basidiomycetes for this method (Table 1). The two basidiomycete proteomes were included for the off chance of analysing gene clusters that do not originate from ascomycetes. The database covers the complete ascomycetes phylum and was iteratively tested and optimized for the detection of co-evolution in ascomycetes. The sequences were downloaded from the National Center for Biotechnology Information (NCBI) database and the Joint Genome Institute (JGI) [44]. A short identifier, unique in the database for each proteome, was introduced to enable multiple pairwise tree comparisons by the treeKO application [45]. A custom Perl script was used for removing duplicated entries in the database. The database is deposited in the GitHub repository https://github.com/gvignolle/FunOrder (doi:10.5281/zenodo.5118984).

thumbnail
Download:
Table 1. Fungal proteomes included in the empirically optimized database.

https://doi.org/10.1371/journal.pcbi.1009372.t001

Workflow

The workflow for the FunOrder method is depicted in Fig 2. First, the sequences of the BGC to be analysed are fed into the software bundle. FunOrder accepts a single file in either genbank file format or fasta format as input. The input files contain BGCs predicted by tools such as antiSMASH [29] or DeepBGC [33]. In case a genbank file is provided, a python script (Genbank to FASTA by Cedar McKay and Gabrielle Rocap, University of Washington) is called to extract the amino acid sequence of the genes in the BGC and create a fasta file. The multi-fasta file is then split into individual fasta files each containing a single protein sequence. These are placed in a subfolder created for the analysis of the BGC. Each file is named either after the position of the gene in the BGC or after the respective protein sequence description. This varies from the input file and the varying annotations used (If needed this can be changed in the script following the instructions of Genbank to FASTA by Cedar McKay and Gabrielle Rocap, University of Washington). Each header of the query sequences is tagged with the identifier "query" at the beginning of the header. The individual sequences are compared to the empirically optimized proteome database (Table 1) by a sequence similarity search using blastp 2.8.1+ (Protein-Protein BLAST) [133]. The output of this search is saved in a file with the ".tab" extension. Additionally, an optional remote search of the non-redundant National Center for Biotechnology Information (NCBI) protein database can be performed, yielding a file with the "ncbi.tab" extension. This allows a preliminary manual analysis of the input sequences and facilitates subsequent annotations of the BGCs.

thumbnail
Download:
Fig 2. Schematic representation of the workflow of FunOrder.

https://doi.org/10.1371/journal.pcbi.1009372.g002

Next, the top 20 results of the blastp analysis are extracted and combined with the query sequence for each gene. A custom Perl script removes potential duplicate entries based on sequence identity. Using emma, a multiple sequence alignment of these protein sequences is calculated based on the ClustalW [134] algorithm, and a dendrogram computed. Based on the multiple sequence alignment, 100 rapid Bootstraps and a subsequent search for the best-scoring maximum likelihood (ML) tree are performed using RAxML (Randomized Axelerated Maximum Likelihood) [135]. The phylogenetic trees are computed using the LG amino acid substitution model. Furthermore, a standard ascertainment bias correction by Paul O. Lewis is performed. At this stage, we have obtained a phylogenetic tree (within the context of our empirically optimized database) for each protein of the input BGC.

To estimate if and to what extent the different genes within a BGC are co-evolved, the strict distance and speciation distance among the ML trees of the individual genes are calculated using the TreeKO algorithm [45]. This tool was designed for automated tree comparison and was already suggested to be used for the detection of co-evolution in protein families [45]. The tool compares the topology of different trees; a distance of 0 in both distance measures represents identical trees. In this context, a higher similarity between the different trees of the individual genes points towards a shared evolution. The strict distance is a weighted Robinson-Foulds (RF) distance measure that penalizes dissimilarities in evolutionarily important events such as gene losses and gene duplications; it has been suggested to be more significant in the detection of co-evolution than the evolutionary distance [45]. In contrast, the evolutionary or speciation distance is computed without taking evolutionary exceptions, such as duplication events or different species content of the two compared trees into account and infers shared "speciation history" based solely on topology without considering branch lengths and only considering shared species of the compared trees. Therefore, an evolutionary distance of 0 does not necessarily describe identical trees but shared "speciation history" of shared species. All pairwise strict and evolutionary distances are combined into matrices which are used as input for an R script [136–140].

In this R-script, first, the strict and evolutionary distances are summed up to a third combined distance matrix combining the information about co-evolution and shared speciation into a single measure. In our experience, this measure can be helpful to detect genes that share little co-evolution with the core-enzymes but are still essential for the biosynthesis, which is reflected in a shared speciation. The evolutionary distance is not directly part of the output of FunOrder as is not intended to be used for the detection of co-evolution. Second, the strict and the combined distance matrices are visualized as heatmaps with a dendrogram computed with the complete linkage method, to find similar clusters in these data sets. Next, the Euclidean distance within the matrices is computed and clustered using Ward’s minimum variance method aiming at finding compact spherical clusters, with the implemented squaring of the dissimilarities before cluster updating, for the two distance matrices separately, with scaled input data [141]. Lastly, a principal component analysis (PCA) is performed on the two distance matrices and the score plot of the first two principal components visualized, respectively. These outputs enable the adoption of a larger view on the distance measures and thereby allow the analysis of co-evolution within the BGC from different perspectives. We describe in a following subchapter how to interpret these visualisations.

The software bundle is written in the BASH (Bourn Again Shell) environment and includes all necessary subprograms. As BASH is the default shell-language of all Linux distributions and MacOS, FunOrder can run on these two operation systems. The FunOrder software package is deposited in the GitHub repository https://github.com/gvignolle/FunOrder (doi:10.5281/zenodo.5118984). Notably, the software package includes scripts adapted to the use on servers and for the integration in various pipelines; details on these can be found in the ReadMe file on the GitHub repository. FunOrder requires some dependencies e.g., RAxML (Randomized Axelerated Maximum Likelihood) [135] and the EMBOSS (The European Molecular Biology Open Software Suite) package [142], for details and links to all dependencies please refer to the ReadMe file on the GitHub repository.

Compilation of benchmark gene clusters (GCs)

To test and evaluate the applicability of the FunOrder method, we used different control and test gene (or protein) sets. The sequences of all test and control sets are deposited in the GitHub repository https://github.com/gvignolle/FunOrder (doi: 10.5281/zenodo.5118984). The first set of negative control gene clusters (GCs) were 42 completely randomly generated synthetic GCs, which were created with a custom BASH script. Therein, ATGC strings of random composition and length were translated to amino acid strings using transeq from the EMBOSS package and the asterisks were removed. The second set of negative controls were 60 random GCs which were created by subsampling randomly the fungal proteome database with a Perl script from the MEME suit [143]. For each random GC a different seed number was given to guarantee non repetitive GCs, each random GC contained 3–10 randomly chosen protein sequences in a random order. These negative control GCs were subsampled from different genomes to maximize the randomness and use gene clusters that should not contain co-evolved genes.

We used a set of 30 empirically well characterized BGCs from a broad range of different genera (Table 2) as positive controls. The BGC sequences were downloaded from NCBI or the MIBiG (Minimum information about a biosynthetic gene cluster) database [144]. The sequences are available at the GitHub repository https://github.com/gvignolle/FunOrder (doi:10.5281/zenodo.5118984). All BGCs were manually inspected for correctness and completeness based on the respective literature (S1 Table, references in Table 2). We further added 2 genes on each side of the BGC to mimic the greedy gain performed by antiSMASH, if possible (sequences available) and applicable (only few or no gap genes present). Next, we defined the class of each gene (biosynthetic gene, further essential gene, gap, or extra gene) according to the described function of the enzymes in the literature (S1 Table).

thumbnail
Download:
Table 2. Empirically characterized biosynthetic gene clusters used as positive controls.

https://doi.org/10.1371/journal.pcbi.1009372.t002

Further, we compiled 10 protein sets containing the sequences of enzymes of conserved metabolic pathways from organisms that were not included in the proteome database, termed „Biosynthetic_pathways“, or „BioPath”(S2 Table; sequences deposited at the GitHub repository https://github.com/gvignolle/FunOrder (doi:10.5281/zenodo.5118984)). As we anticipate a strong co-evolution among the corresponding genes, we used these sets as positive controls for co-evolution in general. Finally, we subsampled the genomes of organisms that were not included in the proteome database for 30 random loci containing 8 to 10 genes (S3 Table; sequences available at the GitHub repository https://github.com/gvignolle/FunOrder (doi:10.5281/zenodo.5118984)). We termed this control set „sequential GCs“. This set should represent the random degree of co-evolution based only on genomic vicinity. Notably, due to the randomness of the sampling, the sequential GCs may also contain evolutionary linked genes.

Calculation of MEM and determination of thresholds for co-evolution

As the thresholds for the strict and/or evolutionary distance for the analysis of protein co-evolution are database dependent, we needed to define these thresholds manually. To this end, we performed a manual comparison of the phylogenetic trees of genes anticipated to be co-evolved and of not presumably co-evolved genes. As positive control datasets (anticipated co-evolution), we used the essential genes within the positive control BGCs. As negative control data set (anticipated to not have co-evolved), we used the genes in the random GCs. For the manual tree comparisons, we considered the topology (defined in S4 Table), branch lengths, number of nodes, and shared leaves of the trees and calculated the manual evaluation measure (MEM) according to the definitions in S5 Table. We calculated the MEM for each gene tree pair of the positive and the negative control data sets (S6 and S7 Tables, respectively). The measure ranges from 3 (same) to 0 (no shared leaves). The MEM values of each pair-wise tree comparison were then manually reconciled with the corresponding strict and the combined distance measures obtained from the treeKO analysis and the subsequent R script, respectively. The procedure is exemplary described for the 2-Pyridon-Desmethylbassianin (dmb) BGC from Beauveria bassiana in S1 File. Based on these manual comparisons, we defined the threshold values for strict and combined distances in the following: two genes are considered as co-evolved if the strict distance value is less than 0.7 or if the combined distance is equal to or less than 60 percent of the maximum value in the combined distance matrix of the analysed set.

Calculation of the Internal co-evolutionary quotient (ICQ)

The internal co-evolutionary quotient (ICQ) expresses how many genes in a GC or proteins in a protein set are co-evolved according to the previously defined threshold for strict and combined distances within the distance matrices of an analysed GC (or protein set). To calculate the ICQ, each protein is compared with every other protein. The total number of all possible pairwise comparisons is 2* [d*(d-1)] for d proteins. The ICQ was calculated using Eq 1, resulting in values between 0 and 1, with 1 representing no co-evolved genes, and 0 representing that most genes are co-evolved with each other in the insert GC.

Eq 1

ICQ = internal co-evolutionary quotient; g = number of strict distances < 0.7 and combined distances < = (0.6 * max value of the combined distance matrix) in all matrices (visualized in the heatmaps); d = number of genes in the GC.

Manual interpretation of the FunOrder output

The FunOrder outputs three different visualizations (heatmap, dendrogram, PCA) each of the strict and combined distance matrices among the genes (or proteins) of an inserted GC (or protein set). These visualizations need to be interpreted manually. For the manual interpretation, we first searched for genes that clustered together with the core enzyme(s) in any of the three visualisations of the strict distance. The definition of the clusters needs to be performed carefully keeping the biological background (gene predictions) in mind. For instance, a cluster containing typical tailoring enzymes (e.g., hydrolases, P450 cytochrome oxidases, FAD-containing enzymes, etc.) and/or further essential genes (e.g., transcription factors or transporters) make sense, whereas clusters containing a lot of genes encoding for unknown genes and/or genes that are unlikely to be involved in the biosynthesis of a secondary metabolite) do not make sense. Next, clustering in the visualizations of the combined distances is considered. As the combined distance also contains information about the speciation history, it may be used to add further genes to the list of “detected genes”. Notably, this needs to be critically evaluated and decided on a case-to-case basis, taking the gene predictions into account. Please also refer to S2 File for a detailed step-by-step description of the interpretation procedure, the exemplary analysis of the lovastatin BGC from A. terreus in the results, and S3 File and S4 File for the exemplary analysis of two unknown BGCs.

Performance evaluation

To test the robustness of FunOrder, we analysed 42 completely randomly generated synthetic GCs. To test whether the FunOrder method can be used to detect co-evolution within GCs (or protein sets), we calculated the ICQ for different control sets and compared the results in a kernel density plot. To evaluate the performance of the FunOrder method regarding its capability to identify presumably co-evolved essential genes (as defined in S1 Table) and to distinguish them from (presumably not co-evolved) gap genes and genes outside of the BGC via the detection of co-evolution, we performed a manual interpretation of 30 empirically characterized BGCs (Table 2) as described above. Genes that clustered together with the core enzyme(s) according to the procedure described above were considered as „detected“. Then we counted the total number of (1a) detected essential genes or (1b) detected biosynthetic genes, (2a) not detected essential genes or (2b) not detected biosynthetic genes, (3) detected gap and extra genes, and (4) not detected gap or extra genes in all BGCs, and defined (1a or 1b) as true positives (TP), (2a or 2b) as false negatives (FN), (3) as false positives (FP), and (4) as true negatives (TN). The values were used for a final statistical evaluation of FunOrder as suggested by Chicco and Jurman [198].

Results and discussion

Applicability of FunOrder for the detection of co-evolution

First, we analyzed the 42 synthetic negative control GCs with the FunOrder software. We could not find any sequence similarities with the empirically optimized fungal proteome database, demonstrating the robustness of the FunOrder method towards non-biological random amino acid sequences. Consequently, the 42 synthetic negative control GCs were not considered in the following.

Next, we performed FunOrder analyses of different control GCs and protein sets and calculated the internal co-evolutionary quotients (ICQs) using Eq 1. The ICQ is a value for the relative amount of co-evolutionary relations among the genes (or proteins) in a given GC or protein set. An ICQ of 0 means that most genes (or proteins) are co-evolved with each other. An ICQ of 1 means, that no co-evolution can be detected using the defined thresholds. As negative control for co-evolution, we used 60 randomly assembled negative control GCs (random GCs, S8 Table). The random GCs were compiled by subsampling different proteomes, to minimize the chance of random, unwanted co-evolution in the clusters. As positive control for co-evolution we used 10 protein sets from conserved metabolic pathways of different ascomycetes (S2 Table), termed „Biosynthetic pathways“, or „BioPath“. Given, that the proteins are part of the conserved primary metabolism and that their enzymatic functions are interrelated, we can assume a high level of internal co-evolution among the proteins within these protein sets. As control for the basic co-evolutionary value of co-localized (or sequential) genes, we used 30 random genetic loci containing 8 to 10 genes (S3 Table). We termed this control set „sequential GCs“. As test set for BGCs of the secondary metabolism in ascomycetes we used 30 empirically characterized BGCs (Table 2, S1 Table), also termed positive control BGCs.

We compared the ICQs of the different sets in an ANOVA (S5 File) and in a kernel density plot (Fig 3). We found that the ICQs for the random GCs were significantly different from all the other sets, demonstrating that the workflow of the FunOrder method can be used to detect co-evolution, that the ICQ is a meaningful measure to represent the content of co-evolutionary relationships within a GC or protein set, and that the manually defined thresholds for strict and combined distances are applicable to define co-evolution within GC or proteins sets. Based on these results, we defined the threshold of the ICQ for biologically relevant co-evolution within a GC as the point of intersect between the random GCs and the BGCs (0.718). GCs with an ICQ above this threshold do not contain significantly more co-evolutionary connections among the contained genes than randomly assembled GCs.

thumbnail
Download:
Fig 3. Kernel density plot of the ICQ values for co-evolutionary linked enzymes of different control sets.

BioPath, protein sets of conserved biosynthetic pathways of the primary metabolism (S2 Table); random GCs, randomly assembled protein sets from 134 fungal proteomes (Table 1); BGCs, previously empirically characterized fungal BGCs (Table 2); sequential GCs, co-localized genes from random loci of different ascomycetes (S3 Table).

https://doi.org/10.1371/journal.pcbi.1009372.g003

To our surprise, we could not detect a statistically significant difference between the sequential GCs and the positive control GCs. However, the maxima for the BioPath proteins and the BGC are at the same value and the shape of the corresponding density plot is remarkably similar (Fig 3), whereas the maximum of the sequential GC is shifted towards the random GCs and the shape of the curve is different to the two positive control sets (Fig 3). These results indicate, that using only the absolute values of strict and combined distance may not be enough to distinguish co-evolutionary linked genes within the context of co-localized genes, but that the distances need to be assessed and interpreted in a case-by-case scenario considering the biological background and context of the analyzed GC.

Exemplary analysis of the lovastatin BGC (lov)

The FunOrder method allows the detection of co-evolved genes within a set of genes or proteins. As mentioned, we speculate that essential genes in BGCs are co-evolving and can therefore be differentiated from gap genes. In this context, the application of FunOrder might be used to detect the essential or at least the biosynthetic genes in BGCs. The software package of the FunOrder method calculates two distance matrices for the proteins within an input GC representing the evolutionary similarities (based on pair-wise comparisons of the phylogenetic trees using the treeKO tool [45]). First, we tried to use the previously defined thresholds for the strict and combined distances to automatically detect the co-evolutionary relations in BGCs. As insinuated above, this proofed not to be a successful strategy (not shown). We speculate, that the evolutionary similarities or distances among neighbouring genes are highly location specific and that the absolute values are therefore not meaningful as general thresholds. However, as the underlying strategy and method is clearly able to detect co-evolution (Fig 3), we speculated that the obtained data may need to be represented in different forms and/or reduced. Consequently, we added the following data visualizations to the FunOrder pipeline. The strict and combined distances are visualized in a heatmap and clustered by higher similarities (complete linkage method). Next, the Euclidean distances within the scaled distance matrices are calculated and clustered (hierarchical clustering) using the Wards minimum variance method aiming at finding compact spherical clusters, with the implemented squaring of dissimilarities before cluster updating. The clustering is visualized in dendrograms. Finally, the principal components of the data are represented in a score plot. Here, we exemplary describe the manual interpretation of these visualizations (S6 File and Fig 4) with the aim to detect co-evolution within the lovastatin BGC of A. terreus (lov, Fig 1). Please refer also to the step-by-step description on how to interpret the FunOrder output in S2 File.

thumbnail
Download:
Fig 4. A selection of the standard output of the FunOrder analysis of the lovastatin BGC (lov).

Score plots of the first two principal components from a PCA performed on the strict distance matrix (A) and on the combined distance matrix (B). The biosynthetic genes and the further essential genes are indicated in red and gold, respectively. Clusters in the PCA are indicated by the dashed boxes.

https://doi.org/10.1371/journal.pcbi.1009372.g004

For the analysis of the lovastatin BGC, we first had a look at the heatmap representing the strict distance matrix (S6 File). Therein all biosynthetic genes (lovA-D, F, G; Fig 1, red arrows) are clustering together with each other and with the gap gene orf1, although not all inter-gene distances were below the previously defined threshold (S6 File, heatmaps). This demonstrates again that, evaluating only the numerical values (regardless of the concrete thresholds) is not enough for a thorough analysis of a BGC. It is necessary to consider the distances within the genomic context by comparing all provided visualisations. The biosynthetic genes of lovastatin (lovA-D, F, G) also formed distinct clusters in the dendrograms and in the PCA of the strict distance (S6 File and Fig 4A) In our experience, it was often helpful to additionally take the combined distance values into consideration to get a more comprehensive picture of the BGC. As mentioned before, the combined distance also considers speciation history. In the case of the lovastatin BGC, orf10 and extra03 clustered together with lovA, B, D, F, G in the PCA of the combined distance (Fig 4B). The gene orf10 encodes for an MFS (major facilitator superfamily) transporter, which warrants adding it to the „detected genes“; the transporter is actually necessary for the export of lovastatin [17] (Fig 1). The gene extra03 is predicted to encode for an alpha-glucuronidase (AguA) which is involved in the hydrolysis of xylan. Therefore, the clustering only in combined distance matrix does not justify classifying the gene extra03 as „detected“. The other two „further essential genes“, lovE and orf8 did not cluster together with the biosynthetic genes in any visualizations of the distance matrices (Fig 4 and S6 File)). LovE is a transcription factor and the main regulator of the lovastatin cluster [17] and essential for the lovastatin biosynthesis in the native organism, although it is not directly part of the biosynthetic pathway. The gene orf8 encodes for a 3-hydroxy-3-methylglutaryl coenzyme-A (HMG-CoA) reductase, which is the target of statins [199] and in this case is conveying self-resistance to lovastatin [200]. These results suggest that these two genes did not undergo the same evolutionary process as the biosynthetic genes. This is in accordance with the „brick and mortar”model suggested by Medema et al. [40]. The biosynthetic genes represent a co-evolving „brick“, that is integrated into the biological context of A. terreus via the „mortar”that are the further essential genes.

This exemplary analysis demonstrates how the different data output formats of the software package need to be considered and compared manually, to decide on which genes are co-evolutionary linked and likely to be involved in the biosynthesis of a secondary metabolite. When considering only one output, one might get a distorted view of the analysed BGC. Notably, we did not intend to leave this step up to automation, because the human (expert or child) pattern recognition and mind still outperforms artificial intelligence (AI) algorithms and machine learning algorithms in this regard [201]. Please also refer to S3 File and S4 File in which we describe the analysis of two yet undescribed BGCs.

Speed and scalability of the software

As the empirically optimized proteome database contained only 134 fungal proteomes, we were able to use the blastp algorithm for sequence similarity search. The analysis of the lovastatin BGC of A. terreus (lov) with 17 genes, took 1 h 19 m 48 sec real time using 22 threads on an Ubuntu Linux system with 128 GB DDR4 RAM. The same analysis took 6 h 54 m 50 sec real time using 3 threads and 5 h 48 m 50 sec using 4 threads on a Linux Mint Laptop, demonstrating that the analysis of such a large cluster as the lovastatin cluster is fast and feasible. The number of threads can be defined, to increase the scalability and the overall performance.

Performance evaluation

Up to this point, we demonstrated that the FunOrder method can be used to detect the overall level of internal co-evolutionary relations within a GC or set of proteins. We demonstrated that similar levels of co-evolutionary relations occur among the genes in BGCs and among proteins of conserved metabolic pathways of the primary metabolism, and that these positive control sets can be distinguished from negative control GC, containing randomly stringed together proteins from different organisms with a threshold of 0.718 for the ICQ (Fig 3). Further, we showed that the values of strict and combined distances need to be visualized in different forms and then interpreted manually to detect co-evolution of individual genes within fungal BGCs. Next, we aimed to test, whether the detection of co-evolved genes is indeed a useful approach to identify the essential genes in fungal BGCs. To this end, we analysed the 30 empirically verified BGCs (Table 2) as described for the lovastatin cluster before. We looked for genes that are co-evolutionary linked with the core biosynthetic gene. These genes were considered as “detected”. The “detected” genes sets were compared to the previously empirically obtained set of essential genes and classified the genes in true positives (TP), false negatives (FN), false positives (FP), or true negatives (TN) (S1 Table). To test and evaluate, how well FunOrder is performing in detecting either all essential or just the biosynthetic genes, we determined two different sets of TP and FN. TPs were either all detected essential genes, or all detected biosynthetic genes. Accordingly, FNs were either all not detected essential genes or all not detected biosynthetic genes (S1 Table). In both cases, FPs were all detected gap and extra genes, and TNs were all not detected gap and extra genes (S1 Table) because it makes biologically no sense to define a „detected”further essential gene as a FP, even when defining detected biosynthetic genes as TP. For an initial performance estimation, we calculated the percentages of detected essential and biosynthetic genes (S1 Table) and compiled them in a kernel density plot (Fig 5). More than 75% of all essential genes and biosynthetic genes were found to be co-evolving using the FunOrder method in 13 and 16 BGCs (out of 30 BGCs), respectively. The curves in the density plot also differ at high percentages; nearly all (above 90%) biosynthetic genes could be detect in more cases than nearly all essential genes. These two observations point in the direction, that especially the biosynthetic genes share a more coherent co-evolutionary history and can thus be identified by looking for co-evolved genes in BGCs. Obviously, not all essential genes in all BGCs are co-evolving and/or can be detected as co-evolved with this method. This is at least partly based on the biological background. Each BGC has a unique evolutionary background and needs to be interpreted individually. The FunOrder method offers additional information about co-evolution for already defined BGCs and may be useful in deciding which genes might be most relevant when studying a BGC.

thumbnail
Download:
Fig 5. Kernel density plots of the relative discovery rate of essential or biosynthetic genes in 30 tested fungal BGCs.

https://doi.org/10.1371/journal.pcbi.1009372.g005

For a stringent statistical evaluation, we calculated the normalized Matthews correlation coefficient (normMCC) and other classical metrics and global metrics (Table 3) as indicated by Chicco and Jurman [198] based on the previously defined TP, FN, FP, and TN (S1 Table). To determine the degree of balance between positive and negative controls we calculated the no-information error rate ni which is best for balanced test sets with the value 0.5. The obtained values of 0.5084 and 0.5444 allowed for the usage and confirmed the validity of the classical metrics such as F1 score and Accuracy. The FunOrder method displays overall high metrics in identifying essential and/or biosynthetic genes in a BGC. Despite the differences between biosynthetic and essential genes in Fig 5, we could not detect strong differences in the overall statistical assessment. FunOrder can be used to detect essential and biosynthetic genes in a BGC based on protein family co-evolution with a accuracy of 0.7215 and 0.743, respectively.

thumbnail
Download:
Table 3. Statistical evaluation of the performance of FunOrder in detecting relevant genes in BGCs.

https://doi.org/10.1371/journal.pcbi.1009372.t003

Concluding remarks

The FunOrder method was created to identify the essential genes in a BGC and distinguish them from gap genes based on the hypothesis that the essential genes are co-evolutionary linked. We evaluated this method and simultaneously tested the underlying hypothesis using different control sets of genes and proteins, respectively. We observed on the one hand that co-evolutionary linkage in fungal BGCs is commonly occurring—especially within the biosynthetic genes, and on the other hand that the FunOrder method can be used to detect the biosynthetic genes within BGCs and to some extent also the further essential genes. We would like to stress that this method is delivering data on co-evolution, that needs to be critically evaluated and interpreted keeping the biological background in mind, and that FunOrder is not to be considered a stand-alone tool but meant to deliver supplementary data about co-evolution within predefined BGCs.

During the testing and evaluation, we encountered several cases of ambiguous results, where the different visualizations clustered different genes together. One way to handle such ambiguous results is to critically assess the results by considering the gene predictions. We further suggest adding and/or removing genes at the edges of the BGC and re-running the analysis. This might change the clustering behaviour and clarify the results. Alternatively, homologous BGCs from other fungi may be analysed by FunOrder and the clustering of the corresponding genes compared to the initial BGC.

The basis but also limitation for the method is the database [42]. Here we used a specific set of proteomes (Table 1) and were thus able to detect co-evolved genes in ascomycetes. Notably, the underlying strategy and workflow of FunOrder can be adapted to analysing genomic regions in other phyla, orders, or even kingdoms by using different databases. In case a larger database is integrated into the software package, alternative search algorithms, such as DIAMOND [202] or HMMER (similarity search using hidden Markov models) [203] might be used instead of blastp to enhance the performance. Nevertheless, each novel database, even if only one single proteome would be introduced in an existing database, will have to be verified and validated.

In this study, we looked for genes that share the same or a similar evolutionary background with the core genes of BGCs and could demonstrate that FunOrder is a fast and powerful method that can support scientists to decide which genes of a BGC are promising study objects. Notably, the application of this method is not limited to fungal BGC. It can be used for any applications where information of a shared co-evolution can contribute to a better understanding. FunOrder with the existing ascomycete database might already be used for a genome wide analysis of co-evolving transcription factors or detection of functionally connected protein-protein interactions [42]. As a future perspective, FunOrder might be even used for the analysis of total proteomes to detect evolutionary linked genes.

Supporting information

S1 Table. Empirically tested BGCs used as control set in this study.

https://doi.org/10.1371/journal.pcbi.1009372.s001

(XLSX)

S2 Table. Protein sets of conserved metabolic pathways of the primary metabolism.

https://doi.org/10.1371/journal.pcbi.1009372.s002

(XLSX)

S3 Table. Sequential GCs used in this study.

https://doi.org/10.1371/journal.pcbi.1009372.s003

(XLS)

S4 Table. Definition of topology.

https://doi.org/10.1371/journal.pcbi.1009372.s004

(PDF)

S5 Table. Parameters used to calculate the manual evaluation measure (MEM).

https://doi.org/10.1371/journal.pcbi.1009372.s005

(PDF)

S6 Table. Calculation of MEM values for positive control BGCs.

https://doi.org/10.1371/journal.pcbi.1009372.s006

(XLSX)

S7 Table. Calculation of MEM values for negative control GCs.

https://doi.org/10.1371/journal.pcbi.1009372.s007

(XLSX)

S8 Table. Random GCs used in this study.

https://doi.org/10.1371/journal.pcbi.1009372.s008

(XLSX)

S1 File. Exemplary MEM analysis of the dmb BGC.

https://doi.org/10.1371/journal.pcbi.1009372.s009

(PDF)

S2 File. Step-by-step explanation for the manual interpretation of the FunOrder output.

https://doi.org/10.1371/journal.pcbi.1009372.s010

(PDF)

S3 File. Exemplary interpretation of the FunOrder output of an unknown fungal BGC 1.

https://doi.org/10.1371/journal.pcbi.1009372.s011

(PDF)

S4 File. Exemplary interpretation of the FunOrder output of an unknown fungal BGC 2.

https://doi.org/10.1371/journal.pcbi.1009372.s012

(PDF)

S5 File. ANOVA for the ICQ values of the control and tests GCs and protein sets, respectively.

https://doi.org/10.1371/journal.pcbi.1009372.s013

(PDF)

S6 File. FunOrder output of the Lovastatin BGC from A. terreus (lov).

https://doi.org/10.1371/journal.pcbi.1009372.s014

(PDF)

Acknowledgments

We thank Dr. Ekaterina Shelest for reviewing a preprint of this study on https://www.biorxiv.org/ and providing us with helpful inputs and suggestions. We thank Matthias Schmal for testing the FunOrder installation and performance on a Linux Mint Laptop.

References

  1. 1. Thirumurugan D, Cholarajan A, Raja SSS, Vijayakumar R. An Introductory Chapter: Secondary Metabolites. In: Vijayakumar R, Raja SSS, editors. Secondary Metabolites—Sources and Applications. London, UK: IntechOpen Limited; 2018.
    • 2. Malik VS. Microbial secondary metabolism. Trends in Biochemical Sciences. 1980;5(3):68–72. https://doi.org/10.1016/0968-0004(80)90071-7.
    • 3. Keller NP, Turner G, Bennett JW. Fungal secondary metabolism—from biochemistry to genomics. Nature Reviews Microbiology. 2005;3(12):937–47. pmid:16322742
    • 4. Alberti F, Foster GD, Bailey AM. Natural products from filamentous fungi and production by heterologous expression. Applied microbiology and biotechnology. 2017;101(2):493–500. Epub 2016/12/15. pmid:27966047; PubMed Central PMCID: PMC5219032.
    • 5. Newman DJ, Cragg GM, Kingston DGI. Chapter 5—Natural Products as Pharmaceuticals and Sources for Lead Structures**Note: This chapter reflects the opinions of the authors, not necessarily those of the US Government. In: Wermuth CG, Aldous D, Raboisson P, Rognan D, editors. The Practice of Medicinal Chemistry (Fourth Edition). San Diego: Academic Press; 2015. p. 101–39.
      • 6. Brakhage AA, Schroeckh V. Fungal secondary metabolites—strategies to activate silent gene clusters. Fungal genetics and biology: FG & B. 2011;48(1):15–22. pmid:20433937.
      • 7. Atanasov AG, Zotchev SB, Dirsch VM, Orhan IE, Banach M, Rollinger JM, et al. Natural products in drug discovery: advances and opportunities. Nature Reviews Drug Discovery. 2021. pmid:33510482
      • 8. Wiemann P, Keller NP. Strategies for mining fungal natural products. Journal of industrial microbiology & biotechnology. 2014;41(2):301–13. pmid:24146366.
      • 9. Kensler TW, Roebuck BD, Wogan GN, Groopman JD. Aflatoxin: a 50-year odyssey of mechanistic and translational toxicology. Toxicol Sci. 2011;120 Suppl 1:S28–48. pmid:20881231; PubMed Central PMCID: PMC3043084.
      • 10. Blount W. Turkey “X” disease. Turkeys. 1961;9(2):52–5.
      • 11. Yu J, Chang PK, Cary JW, Wright M, Bhatnagar D, Cleveland TE, et al. Comparative mapping of aflatoxin pathway gene clusters in Aspergillus parasiticus and Aspergillus flavus. Applied and environmental microbiology. 1995;61(6):2365–71. pmid:7793957; PubMed Central PMCID: PMC167508.
      • 12. Soldatou S, Eldjarn GH, Huerta-Uribe A, Rogers S, Duncan KR. Linking biosynthetic and chemical space to accelerate microbial secondary metabolite discovery. FEMS microbiology letters. 2019;366(13). pmid:31252431
      • 13. Craney A, Ahmed S, Nodwell J. Towards a new science of secondary metabolism. The Journal of antibiotics. 2013;66(7):387–400. pmid:23612726
      • 14. Osbourn A. Secondary metabolic gene clusters: evolutionary toolkits for chemical innovation. Trends in Genetics. 2010;26(10):449–57. pmid:20739089
      • 15. Tran PN, Yen MR, Chiang CY, Lin HC, Chen PY. Detecting and prioritizing biosynthetic gene clusters for bioactive compounds in bacteria and fungi. Appl Microbiol Biotechnol. 2019;103(8):3277–87. Epub 2019/03/13. pmid:30859257; PubMed Central PMCID: PMC6449301.
      • 16. Keller NP. Fungal secondary metabolism: regulation, function and drug discovery. Nat Rev Microbiol. 2019;17(3):167–80. Epub 2018/12/12. pmid:30531948; PubMed Central PMCID: PMC6381595.
      • 17. Mulder KC, Mulinari F, Franco OL, Soares MS, Magalhaes BS, Parachin NS. Lovastatin production: From molecular basis to industrial process optimization. Biotechnol Adv. 2015;33(6 Pt 1):648–65. Epub 2015/04/15. pmid:25868803.
      • 18. Weber G, Schörgendorfer K, Schneider-Scherzer E, Leitner E. The peptide synthetase catalyzing cyclosporine production in Tolypocladium niveum is encoded by a giant 45.8-kilobase open reading frame. Current Genetics. 1994;26:120–5. pmid:8001164
      • 19. van den Berg MA, Westerlaken I, Leeflang C, Kerkman R, Bovenberg RA. Functional characterization of the penicillin biosynthetic gene cluster of Penicillium chrysogenum Wisconsin54-1255. Fungal genetics and biology: FG & B. 2007;44(9):830–44. Epub 2007/06/06. pmid:17548217.
      • 20. Nosanchuk JD, Stark RE, Casadevall A. Fungal Melanin: What do We Know About Structure? Front Microbiol. 2015;6:1463. Epub 2016/01/07. pmid:26733993; PubMed Central PMCID: PMC4687393.
      • 21. Wheeler MH, Bell AA. Melanins and their importance in pathogenic fungi. Curr Top Med Mycol. 1988;2:338–87. pmid:3288360.
      • 22. Luo S, Dong SH. Recent Advances in the Discovery and Biosynthetic Study of Eukaryotic RiPP Natural Products. Molecules. 2019;24(8). Epub 2019/04/21. pmid:31003555; PubMed Central PMCID: PMC6514808.
      • 23. Montalban-Lopez M, Scott TA, Ramesh S, Rahman IR, van Heel AJ, Viel JH, et al. New developments in RiPP discovery, enzymology and engineering. Nat Prod Rep. 2020. Epub 2020/09/17. pmid:32935693.
      • 24. Wang DN, Toyotome T, Muraosa Y, Watanabe A, Wuren T, Bunsupa S, et al. GliA in Aspergillus fumigatus is required for its tolerance to gliotoxin and affects the amount of extracellular and intracellular gliotoxin. Medical mycology. 2014;52(5):506–18. Epub 2014/05/23. pmid:24847038.
      • 25. Derntl C, Rassinger A, Srebotnik E, Mach RL, Mach-Aigner AR. Identification of the Main Regulator Responsible for Synthesis of the Typical Yellow Pigment Produced by Trichoderma reesei. Applied and environmental microbiology. 2016;82(20):6247–57. pmid:27520818.
      • 26. Schrettl M, Carberry S, Kavanagh K, Haas H, Jones GW, O’Brien J, et al. Self-protection against gliotoxin—a component of the gliotoxin biosynthetic cluster, GliT, completely protects Aspergillus fumigatus against exogenous gliotoxin. PLoS pathogens. 2010;6(6):e1000952. pmid:20548963; PubMed Central PMCID: PMC2883607.
      • 27. Anyaogu DC, Mortensen UH. Heterologous production of fungal secondary metabolites in Aspergilli. Frontiers in microbiology. 2015;6(77). pmid:25713568
      • 28. Chang PK, Yu J, Yu JH. aflT, a MFS transporter-encoding gene located in the aflatoxin gene cluster, does not have a significant role in aflatoxin secretion. Fungal genetics and biology: FG & B. 2004;41(10):911–20. Epub 2004/09/03. pmid:15341913.
      • 29. Blin K, Shaw S, Steinke K, Villebro R, Ziemert N, Lee SY, et al. antiSMASH 5.0: updates to the secondary metabolite genome mining pipeline. Nucleic Acids Res. 2019;47(W1):W81–w7. Epub 2019/04/30. pmid:31032519; PubMed Central PMCID: PMC6602434.
      • 30. Wolf T, Shelest V, Nath N, Shelest E. CASSIS and SMIPS: promoter-based prediction of secondary metabolite gene clusters in eukaryotic genomes. Bioinformatics. 2016;32(8):1138–43. Epub 2015/12/15. pmid:26656005; PubMed Central PMCID: PMC4824125.
      • 31. Khaldi N, Seifuddin FT, Turner G, Haft D, Nierman WC, Wolfe KH, et al. SMURF: Genomic mapping of fungal secondary metabolite clusters. Fungal genetics and biology: FG & B. 2010;47(9):736–41. pmid:20554054; PubMed Central PMCID: PMC2916752.
      • 32. Almeida H, Palys S, Tsang A, Diallo AB. TOUCAN: a framework for fungal biosynthetic gene cluster discovery. NAR Genomics and Bioinformatics. 2020;2(4). pmid:33575642
      • 33. Hannigan GD, Prihoda D, Palicka A, Soukup J, Klempir O, Rampula L, et al. A deep learning genome-mining strategy for biosynthetic gene cluster prediction. Nucleic acids research. 2019;47(18):e110. pmid:31400112; PubMed Central PMCID: PMC6765103.
      • 34. Tai Y, Liu C, Yu S, Yang H, Sun J, Guo C, et al. Gene co-expression network analysis reveals coordinated regulation of three characteristic secondary biosynthetic pathways in tea plant (Camellia sinensis). BMC Genomics. 2018;19(1):616. Epub 2018/08/17. pmid:30111282; PubMed Central PMCID: PMC6094456.
      • 35. Lind AL, Wisecaver JH, Lameiras C, Wiemann P, Palmer JM, Keller NP, et al. Drivers of genetic diversity in secondary metabolic gene clusters within a fungal species. PLOS Biology. 2017;15(11):e2003583. pmid:29149178
      • 36. Rokas A, Wisecaver JH, Lind AL. The birth, evolution and death of metabolic gene clusters in fungi. Nature reviews Microbiology. 2018;16(12):731–44. Epub 2018/09/09. pmid:30194403.
      • 37. Palmer JM, Keller NP. Secondary metabolism in fungi: does chromosomal location matter? Current opinion in microbiology. 2010;13(4):431–6. Epub 2010/07/16. pmid:20627806; PubMed Central PMCID: PMC2922032.
      • 38. Hoogendoorn K, Barra L, Waalwijk C, Dickschat JS, van der Lee TAJ, Medema MH. Evolution and Diversity of Biosynthetic Gene Clusters in Fusarium. Frontiers in microbiology. 2018;9:1158. Epub 2018/06/21. pmid:29922257; PubMed Central PMCID: PMC5996196.
      • 39. Fischbach MA, Walsh CT, Clardy J. The evolution of gene collectives: How natural selection drives chemical innovation. Proceedings of the National Academy of Sciences. 2008;105(12):4601–8. pmid:18216259
      • 40. Medema MH, Cimermancic P, Sali A, Takano E, Fischbach MA. A Systematic Computational Analysis of Biosynthetic Gene Cluster Evolution: Lessons for Engineering Biosynthesis. PLOS Computational Biology. 2014;10(12):e1004016. pmid:25474254
      • 41. Rokas A, Mead ME, Steenwyk JL, Raja HA, Oberlies NH. Biosynthetic gene clusters and the evolution of fungal chemodiversity. Natural product reports. 2020;37(7):868–78. pmid:31898704
      • 42. Ochoa D, Pazos F. Practical aspects of protein co-evolution. Frontiers in Cell and Developmental Biology. 2014;2(14). pmid:25364721
      • 43. Del Carratore F, Zych K, Cummings M, Takano E, Medema MH, Breitling R. Computational identification of co-evolving multi-gene modules in microbial biosynthetic gene clusters. Communications Biology. 2019;2(1):83. pmid:30854475
      • 44. Nordberg H, Cantor M, Dusheyko S, Hua S, Poliakov A, Shabalov I, et al. The genome portal of the Department of Energy Joint Genome Institute: 2014 updates. Nucleic Acids Res. 2014;42(Database issue):D26–31. Epub 2013/11/15. pmid:24225321; PubMed Central PMCID: PMC3965075.
      • 45. Marcet-Houben M, Gabaldon T. TreeKO: a duplication-aware algorithm for the comparison of phylogenetic trees. Nucleic Acids Res. 2011;39(10):e66. Epub 2011/02/22. pmid:21335609; PubMed Central PMCID: PMC3105381.
      • 46. Terfehr D, Dahlmann TA, Specht T, Zadra I, Kurnsteiner H, Kuck U. Genome Sequence and Annotation of Acremonium chrysogenum, Producer of the beta-Lactam Antibiotic Cephalosporin C. Genome announcements. 2014;2(5). pmid:25291769; PubMed Central PMCID: PMC4175204.
      • 47. Nguyen HD, Lewis CT, Lévesque CA, Gräfenhan T. Draft Genome Sequence of Alternaria alternata ATCC 34957. Genome announcements. 2016;4(1). Epub 2016/01/16. pmid:26769939; PubMed Central PMCID: PMC4714121.
      • 48. Hu J, Chen C, Peever T, Dang H, Lawrence C, Mitchell T. Genomic characterization of the conditionally dispensable chromosome in Alternaria arborescens provides evidence for horizontal gene transfer. BMC Genomics. 2012;13(1):171. pmid:22559316
      • 49. Armitage AD, Cockerton HM, Sreenivasaprasad S, Woodhall J, Lane CR, Harrison RJ, et al. Genomics Evolutionary History and Diagnostics of the Alternaria alternata Species Group Including Apple and Asian Pear Pathotypes. Frontiers in microbiology. 2020;10(3124). pmid:32038562
      • 50. Lu Y, Ye C, Che J, Xu X, Shao D, Jiang C, et al. Genomic sequencing, genome-scale metabolic network reconstruction, and in silico flux analysis of the grape endophytic fungus Alternaria sp. MG1. Microb Cell Fact. 2019;18(1):13. Epub 2019/01/27. pmid:30678677; PubMed Central PMCID: PMC6345013.
      • 51. Kohler A, Kuo A, Nagy LG, Morin E, Barry KW, Buscot F, et al. Convergent losses of decay mechanisms and rapid turnover of symbiosis genes in mycorrhizal mutualists. Nat Genet. 2015;47(4):410–5. Epub 2015/02/24. pmid:25706625.
      • 52. Martino E, Morin E, Grelet GA, Kuo A, Kohler A, Daghino S, et al. Comparative genomics and transcriptomics depict ericoid mycorrhizal fungi as versatile saprotrophs and plant mutualists. New Phytol. 2018;217(3):1213–29. Epub 2018/01/10. pmid:29315638.
      • 53. Yang J, Wang L, Ji X, Feng Y, Li X, Zou C, et al. Genomic and proteomic analyses of the fungus Arthrobotrys oligospora provide insights into nematode-trap formation. PLoS Pathog. 2011;7(9):e1002179. Epub 2011/09/13. pmid:21909256; PubMed Central PMCID: PMC3164635.
      • 54. Burmester A, Shelest E, Glöckner G, Heddergott C, Schindler S, Staib P, et al. Comparative and functional genomics provide insights into the pathogenicity of dermatophytic fungi. Genome Biol. 2011;12(1):R7. Epub 2011/01/21. pmid:21247460; PubMed Central PMCID: PMC3091305.
      • 55. Murat C, Payen T, Noel B, Kuo A, Morin E, Chen J, et al. Pezizomycetes genomes reveal the molecular basis of ectomycorrhizal truffle lifestyle. Nat Ecol Evol. 2018;2(12):1956–65. Epub 2018/11/14. pmid:30420746.
      • 56. de Vries RP, Riley R, Wiebenga A, Aguilar-Osorio G, Amillis S, Uchima CA, et al. Comparative genomics reveals high biological diversity and specific adaptations in the industrially and medically important fungal genus Aspergillus. Genome Biology. 2017;18(1):28. pmid:28196534
      • 57. Nierman WC, Yu J, Fedorova-Abrams ND, Losada L, Cleveland TE, Bhatnagar D, et al. Genome Sequence of Aspergillus flavus NRRL 3357, a Strain That Causes Aflatoxin Contamination of Food and Feed. Genome announcements. 2015;3(2). pmid:25883274; PubMed Central PMCID: PMC4400417.
      • 58. Nierman WC, Pain A, Anderson MJ, Wortman JR, Kim HS, Arroyo J, et al. Genomic sequence of the pathogenic and allergenic filamentous fungus Aspergillus fumigatus. Nature. 2005;438(7071):1151–6. Epub 2005/12/24. pmid:16372009.
      • 59. Pel HJ, de Winde JH, Archer DB, Dyer PS, Hofmann G, Schaap PJ, et al. Genome sequencing and analysis of the versatile cell factory Aspergillus niger CBS 513.88. Nature biotechnology. 2007;25(2):221–31. pmid:17259976.
      • 60. Zhao G, Yao Y, Qi W, Wang C, Hou L, Zeng B, et al. Draft genome sequence of Aspergillus oryzae strain 3.042. Eukaryotic cell. 2012;11(9):1178–. pmid:22933657.
      • 61. Vesth TC, Nybo JL, Theobald S, Frisvad JC, Larsen TO, Nielsen KF, et al. Investigation of inter- and intraspecies variation through genome sequencing of Aspergillus section Nigri. Nature Genetics. 2018;50(12):1688–95. pmid:30349117
      • 62. Savitha J, Bhargavi SD, Praveen VK. Complete Genome Sequence of Soil Fungus Aspergillus terreus (KM017963), a Potent Lovastatin Producer. Genome Announc. 2016;4(3). Epub 2016/06/11. pmid:27284150; PubMed Central PMCID: PMC4901219.
      • 63. Spanu PD, Abbott JC, Amselem J, Burgis TA, Soanes DM, Stüber K, et al. Genome expansion and gene loss in powdery mildew fungi reveal tradeoffs in extreme parasitism. Science. 2010;330(6010):1543–6. Epub 2010/12/15. pmid:21148392.
      • 64. Marsberg A, Kemler M, Jami F, Nagel JH, Postma-Smidt A, Naidoo S, et al. Botryosphaeria dothidea: a latent pathogen of global importance to woody plant health. Mol Plant Pathol. 2017;18(4):477–88. Epub 2016/09/30. pmid:27682468; PubMed Central PMCID: PMC6638292.
      • 65. Van Kan JA, Stassen JH, Mosbach A, Van Der Lee TA, Faino L, Farmer AD, et al. A gapless genome sequence of the fungus Botrytis cinerea. Mol Plant Pathol. 2017;18(1):75–89. Epub 2016/02/26. pmid:26913498; PubMed Central PMCID: PMC6638203.
      • 66. Valero-Jiménez CA, Veloso J, Staats M, van Kan JAL. Comparative genomics of plant pathogenic Botrytis species with distinct host specificity. BMC Genomics. 2019;20(1):203. pmid:30866801
      • 67. Knapp DG, Németh JB, Barry K, Hainaut M, Henrissat B, Johnson J, et al. Comparative genomics provides insights into the lifestyle and reveals functional heterogeneity of dark septate endophytic fungi. Sci Rep. 2018;8(1):6321. Epub 2018/04/22. pmid:29679020; PubMed Central PMCID: PMC5910433.
      • 68. Teixeira MM, Moreno LF, Stielow BJ, Muszewska A, Hainaut M, Gonzaga L, et al. Exploring the genomic diversity of black yeasts and relatives (Chaetothyriales, Ascomycota). Stud Mycol. 2017;86:1–28. Epub 2017/03/30. pmid:28348446; PubMed Central PMCID: PMC5358931.
      • 69. Cuomo CA, Untereiner WA, Ma LJ, Grabherr M, Birren BW. Draft Genome Sequence of the Cellulolytic Fungus Chaetomium globosum. Genome announcements. 2015;3(1). pmid:25720678; PubMed Central PMCID: PMC4342419.
      • 70. Armaleo D, Müller O, Lutzoni F, Andrésson Ó S, Blanc G, Bode HB, et al. The lichen symbiosis re-viewed through the genomes of Cladonia grayi and its algal partner Asterochloris glomerata. BMC Genomics. 2019;20(1):605. Epub 2019/07/25. pmid:31337355; PubMed Central PMCID: PMC6652019.
      • 71. Ohm RA, Feau N, Henrissat B, Schoch CL, Horwitz BA, Barry KW, et al. Diverse lifestyles and strategies of plant pathogenesis encoded in the genomes of eighteen Dothideomycetes fungi. PLoS pathogens. 2012;8(12):e1003037. pmid:23236275; PubMed Central PMCID: PMC3516569.
      • 72. Condon BJ, Leng Y, Wu D, Bushley KE, Ohm RA, Otillar R, et al. Comparative genome structure, secondary metabolite, and effector coding capacity across Cochliobolus pathogens. PLoS Genet. 2013;9(1):e1003233. Epub 2013/01/30. pmid:23357949; PubMed Central PMCID: PMC3554632.
      • 73. Baroncelli R, Amby DB, Zapparata A, Sarrocco S, Vannacci G, Le Floch G, et al. Gene family expansions and contractions are associated with host range in plant pathogens of the genus Colletotrichum. BMC Genomics. 2016;17:555. Epub 2016/08/09. pmid:27496087; PubMed Central PMCID: PMC4974774.
      • 74. Baroncelli R, Sanz-Martin JM, Rech GE, Sukno SA, Thon MR. Draft Genome Sequence of Colletotrichum sublineola, a Destructive Pathogen of Cultivated Sorghum. Genome announcements. 2014;2(3). pmid:24926053; PubMed Central PMCID: PMC4056296.
      • 75. Hacquard S, Kracher B, Hiruma K, Münch PC, Garrido-Oter R, Thon MR, et al. Survival trade-offs in plant roots during colonization by closely related beneficial and pathogenic fungi. Nat Commun. 2016;7:11362. Epub 2016/05/07. pmid:27150427; PubMed Central PMCID: PMC4859067.
      • 76. Lopez D, Ribeiro S, Label P, Fumanal B, Venisse JS, Kohler A, et al. Genome-Wide Analysis of Corynespora cassiicola Leaf Fall Disease Putative Effectors. Front Microbiol. 2018;9:276. Epub 2018/03/20. pmid:29551995; PubMed Central PMCID: PMC5840194.
      • 77. Wu W, Davis RW, Tran-Gyamfi MB, Kuo A, LaButti K, Mihaltcheva S, et al. Characterization of four endophytic fungi as potential consolidated bioprocessing hosts for conversion of lignocellulose into advanced biofuels. Appl Microbiol Biotechnol. 2017;101(6):2603–18. Epub 2017/01/13. pmid:28078400.
      • 78. Morales-Cruz A, Amrine KC, Blanco-Ulate B, Lawrence DP, Travadon R, Rolshausen PE, et al. Distinctive expansion of gene families associated with plant cell wall degradation, secondary metabolism, and nutrient uptake in the genomes of grapevine trunk pathogens. BMC Genomics. 2015;16(1):469. Epub 2015/06/19. pmid:26084502; PubMed Central PMCID: PMC4472170.
      • 79. Jones L, Riaz S, Morales-Cruz A, Amrine KC, McGuire B, Gubler WD, et al. Adaptive genomic structural variation in the grape powdery mildew pathogen, Erysiphe necator. BMC Genomics. 2014;15(1):1081. Epub 2014/12/10. pmid:25487071; PubMed Central PMCID: PMC4298948.
      • 80. Blanco-Ulate B, Rolshausen PE, Cantu D. Draft Genome Sequence of the Grapevine Dieback Fungus Eutypa lata UCR-EL1. Genome announcements. 2013;1(3). Epub 2013/06/01. pmid:23723393; PubMed Central PMCID: PMC3668001.
      • 81. Bombassaro A, de Hoog S, Weiss VA, Souza EM, Leão AC, Costa FF, et al. Draft Genome Sequence of Fonsecaea monophora Strain CBS 269.37, an Agent of Human Chromoblastomycosis. Genome Announc. 2016;4(4). Epub 2016/07/30. pmid:27469960; PubMed Central PMCID: PMC4966464.
      • 82. Bashyal BM, Rawat K, Sharma S, Kulshreshtha D, Gopala Krishnan S, Singh AK, et al. Whole Genome Sequencing of Fusarium fujikuroi Provides Insight into the Role of Secretory Proteins and Cell Wall Degrading Enzymes in Causing Bakanae Disease of Rice. Front Plant Sci. 2017;8:2013-. pmid:29230233.
      • 83. Cuomo CA, Guldener U, Xu JR, Trail F, Turgeon BG, Di Pietro A, et al. The Fusarium graminearum genome reveals a link between localized polymorphism and pathogen specialization. Science. 2007;317(5843):1400–2. Epub 2007/09/08. 317/5843/1400 [pii] pmid:17823352.
      • 84. Ma LJ, van der Does HC, Borkovich KA, Coleman JJ, Daboussi MJ, Di Pietro A, et al. Comparative genomics reveals mobile pathogenicity chromosomes in Fusarium. Nature. 2010;464(7287):367–73. Epub 2010/03/20. pmid:20237561; PubMed Central PMCID: PMC3048781.
      • 85. Niehaus EM, Münsterkötter M, Proctor RH, Brown DW, Sharon A, Idan Y, et al. Comparative "Omics" of the Fusarium fujikuroi Species Complex Highlights Differences in Genetic Potential and Metabolite Synthesis. Genome Biol Evol. 2016;8(11):3574–99. Epub 2017/01/04. pmid:28040774; PubMed Central PMCID: PMC5203792.
      • 86. Gardiner DM, Benfield AH, Stiller J, Stephen S, Aitken K, Liu C, et al. A high-resolution genetic map of the cereal crown rot pathogen Fusarium pseudograminearum provides a near-complete genome assembly. Mol Plant Pathol. 2018;19(1):217–26. Epub 2016/11/27. pmid:27888554; PubMed Central PMCID: PMC6638115.
      • 87. Okagaki LH, Nunes CC, Sailsbery J, Clay B, Brown D, John T, et al. Genome Sequences of Three Phytopathogenic Species of the Magnaporthaceae Family of Fungi. G3 (Bethesda). 2015;5(12):2539–45. Epub 2015/09/30. pmid:26416668; PubMed Central PMCID: PMC4683626.
      • 88. Peter M, Kohler A, Ohm RA, Kuo A, Krützmann J, Morin E, et al. Ectomycorrhizal ecology is imprinted in the genome of the dominant symbiotic fungus Cenococcum geophilum. Nat Commun. 2016;7:12662. Epub 2016/09/08. pmid:27601008; PubMed Central PMCID: PMC5023957.
      • 89. Dean RA, Talbot NJ, Ebbole DJ, Farman ML, Mitchell TK, Orbach MJ, et al. The genome sequence of the rice blast fungus Magnaporthe grisea. Nature. 2005;434(7036):980–6. Epub 2005/04/23. pmid:15846337.
      • 90. Gao Q, Jin K, Ying SH, Zhang Y, Xiao G, Shang Y, et al. Genome sequencing and comparative transcriptomics of the model entomopathogenic fungi Metarhizium anisopliae and M. acridum. PLoS Genet. 2011;7(1):e1001264. Epub 2011/01/22. pmid:21253567; PubMed Central PMCID: PMC3017113.
      • 91. Hu X, Xiao G, Zheng P, Shang Y, Su Y, Zhang X, et al. Trajectory and genomic determinants of fungal-pathogen speciation and host adaptation. Proc Natl Acad Sci U S A. 2014;111(47):16796–801. Epub 2014/11/05. pmid:25368161; PubMed Central PMCID: PMC4250126.
      • 92. Binneck E, Lastra CCL, Sosa-Gómez DR. Genome Sequence of Metarhizium rileyi, a Microbial Control Agent for Lepidoptera. Microbiology Resource Announcements. 2019;8(36):e00897–19. pmid:31488537
      • 93. Meerupati T, Andersson K-M, Friman E, Kumar D, Tunlid A, Ahrén D. Genomic Mechanisms Accounting for the Adaptation to Parasitism in Nematode-Trapping Fungi. PLOS Genetics. 2013;9(11):e1003909. pmid:24244185
      • 94. Tan H, Kohler A, Miao R, Liu T, Zhang Q, Zhang B, et al. Multi-omic analyses of exogenous nutrient bag decomposition by the black morel Morchella importuna reveal sustained carbon acquisition and transferring. Environ Microbiol. 2019;21(10):3909–26. Epub 2019/07/18. pmid:31314937.
      • 95. Coleman JJ, Rounsley SD, Rodriguez-Carres M, Kuo A, Wasmann CC, Grimwood J, et al. The genome of Nectria haematococca: contribution of supernumerary chromosomes to gene expansion. PLoS Genet. 2009;5(8):e1000618. Epub 2009/08/29. pmid:19714214; PubMed Central PMCID: PMC2725324.
      • 96. Galagan JE, Calvo SE, Borkovich KA, Selker EU, Read ND, Jaffe D, et al. The genome sequence of the filamentous fungus Neurospora crassa. Nature. 2003;422(6934):859–68. Epub 2003/04/25. pmid:12712197.
      • 97. Baker SE, Schackwitz W, Lipzen A, Martin J, Haridas S, LaButti K, et al. Draft Genome Sequence of Neurospora crassa Strain FGSC 73. Genome announcements. 2015;3(2):e00074–15. pmid:25838471.
      • 98. Ellison CE, Stajich JE, Jacobson DJ, Natvig DO, Lapidus A, Foster B, et al. Massive changes in genome architecture accompany the transition to self-fertility in the filamentous fungus Neurospora tetrasperma. Genetics. 2011;189(1):55–69. Epub 2011/07/14. pmid:21750257; PubMed Central PMCID: PMC3176108.
      • 99. Haridas S, Wang Y, Lim L, Massoumi Alamouti S, Jackman S, Docking R, et al. The genome and transcriptome of the pine saprophyte Ophiostoma piceae, and a comparison with the bark beetle-associated pine pathogen Grosmannia clavigera. BMC Genomics. 2013;14:373. Epub 2013/06/04. pmid:23725015; PubMed Central PMCID: PMC3680317.
      • 100. Urquhart AS, Mondo SJ, Mäkelä MR, Hane JK, Wiebenga A, He G, et al. Genomic and Genetic Insights Into a Cosmopolitan Fungus, Paecilomyces variotii (Eurotiales). Front Microbiol. 2018;9:3058. Epub 2019/01/09. pmid:30619145; PubMed Central PMCID: PMC6300479.
      • 101. Reynolds HT, Vijayakumar V, Gluck-Thaler E, Korotkin HB, Matheny PB, Slot JC. Horizontal gene cluster transfer increased hallucinogenic mushroom diversity. Evolution Letters. 2018;2(2):88–101. pmid:30283667
      • 102. Desjardins CA, Champion MD, Holder JW, Muszewska A, Goldberg J, Bailão AM, et al. Comparative genomic analysis of human fungal pathogens causing paracoccidioidomycosis. PLoS Genet. 2011;7(10):e1002345. Epub 2011/11/03. pmid:22046142; PubMed Central PMCID: PMC3203195.
      • 103. Cheeseman K, Ropars J, Renault P, Dupont J, Gouzy J, Branca A, et al. Multiple recent horizontal transfers of a large genomic region in cheese making fungi. Nat Commun. 2014;5:2876. Epub 2014/01/11. PubMed Central PMCID: PMC3896755. pmid:24407037
      • 104. Specht T, Dahlmann TA, Zadra I, Kürnsteiner H, Kück U. Complete Sequencing and Chromosome-Scale Genome Assembly of the Industrial Progenitor Strain P2niaD18 from the Penicillin Producer Penicillium chrysogenum. Genome announcements. 2014;2(4). Epub 2014/07/26. pmid:25059858; PubMed Central PMCID: PMC4110216.
      • 105. Marcet-Houben M, Ballester A-R, de la Fuente B, Harries E, Marcos JF, González-Candelas L, et al. Genome sequence of the necrotrophic fungus Penicillium digitatum, the main postharvest pathogen of citrus. BMC Genomics. 2012;13(1):646. pmid:23171342
      • 106. Ballester AR, Marcet-Houben M, Levin E, Sela N, Selma-Lázaro C, Carmona L, et al. Genome, Transcriptome, and Functional Analyses of Penicillium expansum Provide New Insights Into Secondary Metabolism and Pathogenicity. Mol Plant Microbe Interact. 2015;28(3):232–48. Epub 2014/10/23. pmid:25338147.
      • 107. Nielsen JC, Grijseels S, Prigent S, Ji B, Dainat J, Nielsen KF, et al. Global analysis of biosynthetic gene clusters reveals vast potential of secondary metabolite production in Penicillium species. Nat Microbiol. 2017;2:17044. Epub 2017/04/04. pmid:28368369.
      • 108. Liu G, Zhang L, Wei X, Zou G, Qin Y, Ma L, et al. Genomic and Secretomic Analyses Reveal Unique Features of the Lignocellulolytic Enzyme System of Penicillium decumbens. PloS one. 2013;8(2):e55185. pmid:23383313
      • 109. van den Berg MA, Albang R, Albermann K, Badger JH, Daran JM, Driessen AJ, et al. Genome sequencing and analysis of the filamentous fungus Penicillium chrysogenum. Nature biotechnology. 2008;26(10):1161–8. pmid:18820685.
      • 110. Wang X, Zhang X, Liu L, Xiang M, Wang W, Sun X, et al. Genomic and transcriptomic analysis of the endophytic fungus Pestalotiopsis fici reveals its lifestyle and high potential for synthesis of natural products. BMC Genomics. 2015;16(1):28. Epub 2015/01/28. pmid:25623211; PubMed Central PMCID: PMC4320822.
      • 111. Blanco-Ulate B, Rolshausen P, Cantu D. Draft Genome Sequence of the Ascomycete Phaeoacremonium aleophilum Strain UCR-PA7, a Causal Agent of the Esca Disease Complex in Grapevines. Genome announcements. 2013;1(3). Epub 2013/07/03. pmid:23814032; PubMed Central PMCID: PMC3695428.
      • 112. Walker AK, Frasz SL, Seifert KA, Miller JD, Mondo SJ, LaButti K, et al. Full Genome of Phialocephala scopiformis DAOMC 229536, a Fungal Endophyte of Spruce Producing the Potent Anti-Insectan Compound Rugulosin. Genome announcements. 2016;4(1). Epub 2016/03/08. pmid:26950333; PubMed Central PMCID: PMC4767923.
      • 113. Cissé OH, Pagni M, Hauser PM. De novo assembly of the Pneumocystis jirovecii genome from a single bronchoalveolar lavage fluid specimen from a patient. mBio. 2012;4(1):e00428–12. Epub 2012/12/28. pmid:23269827; PubMed Central PMCID: PMC3531804.
      • 114. Chibucos MC, Crabtree J, Nagaraj S, Chaturvedi S, Chaturvedi V. Draft Genome Sequences of Human Pathogenic Fungus Geomyces pannorum Sensu Lato and Bat White Nose Syndrome Pathogen Geomyces (Pseudogymnoascus) destructans. Genome announcements. 2013;1(6). Epub 2013/12/21. pmid:24356829; PubMed Central PMCID: PMC3868853.
      • 115. Mondo SJ, Dannebaum RO, Kuo RC, Louie KB, Bewick AJ, LaButti K, et al. Widespread adenine N6-methylation of active genes in fungi. Nat Genet. 2017;49(6):964–8. Epub 2017/05/10. pmid:28481340.
      • 116. Cubeta MA, Thomas E, Dean RA, Jabaji S, Neate SM, Tavantzis S, et al. Draft Genome Sequence of the Plant-Pathogenic Soil Fungus Rhizoctonia solani Anastomosis Group 3 Strain Rhs1AP. Genome Announc. 2014;2(5). Epub 2014/11/02. pmid:25359908; PubMed Central PMCID: PMC4214984.
      • 117. Liti G, Ba ANN, Blythe M, Müller CA, Bergström A, Cubillos FA, et al. High quality de novo sequencing and assembly of the Saccharomyces arboricolus genome. BMC Genomics. 2013;14(1):69. pmid:23368932
      • 118. Goffeau A, Barrell BG, Bussey H, Davis RW, Dujon B, Feldmann H, et al. Life with 6000 genes. Science. 1996;274(5287):546, 63–7. Epub 1996/10/25. pmid:8849441.
      • 119. Quandt CA, Bushley KE, Spatafora JW. The genome of the truffle-parasite Tolypocladium ophioglossoides and the evolution of antifungal peptaibiotics. BMC Genomics. 2015;16(1):553. Epub 2015/07/29. pmid:26215153; PubMed Central PMCID: PMC4517408.
      • 120. Quandt CA, Patterson W, Spatafora JW. Harnessing the power of phylogenomics to disentangle the directionality and signatures of interkingdom host jumping in the parasitic fungal genus Tolypocladium. Mycologia. 2018;110(1):104–17. Epub 2018/06/05. pmid:29863984.
      • 121. Proctor RH, McCormick SP, Kim HS, Cardoza RE, Stanley AM, Lindo L, et al. Evolution of structural diversity of trichothecenes, a family of toxins produced by plant pathogenic and entomopathogenic fungi. PLoS Pathog. 2018;14(4):e1006946. Epub 2018/04/13. pmid:29649280; PubMed Central PMCID: PMC5897003.
      • 122. Druzhinina IS, Chenthamara K, Zhang J, Atanasova L, Yang D, Miao Y, et al. Massive lateral transfer of genes encoding plant cell wall-degrading enzymes to the mycoparasitic fungus Trichoderma from its plant-associated hosts. PLoS Genet. 2018;14(4):e1007322. Epub 2018/04/10. pmid:29630596; PubMed Central PMCID: PMC5908196.
      • 123. Kubicek CP, Herrera-Estrella A, Seidl-Seiboth V, Martinez DA, Druzhinina IS, Thon M, et al. Comparative genome sequence analysis underscores mycoparasitism as the ancestral life style of Trichoderma. Genome Biol. 2011;12(4):R40. Epub 2011/04/20. pmid:21501500; PubMed Central PMCID: PMC3218866.
      • 124. Baroncelli R, Piaggeschi G, Fiorini L, Bertolini E, Zapparata A, Pè ME, et al. Draft Whole-Genome Sequence of the Biocontrol Agent Trichoderma harzianum T6776. Genome announcements. 2015;3(3):e00647–15. pmid:26067977.
      • 125. Xie BB, Qin QL, Shi M, Chen LL, Shu YL, Luo Y, et al. Comparative genomics provide insights into evolution of trichoderma nutrition style. Genome Biol Evol. 2014;6(2):379–90. Epub 2014/02/01. pmid:24482532; PubMed Central PMCID: PMC3942035.
      • 126. Martinez D, Berka RM, Henrissat B, Saloheimo M, Arvas M, Baker SE, et al. Genome sequencing and analysis of the biomass-degrading fungus Trichoderma reesei (syn. Hypocrea jecorina). Nature biotechnology. 2008;26(5):553–60. Epub 2008/05/06. pmid:18454138.
      • 127. Martinez DA, Oliver BG, Gräser Y, Goldberg JM, Li W, Martinez-Rossi NM, et al. Comparative genome analysis of Trichophyton rubrum and related dermatophytes reveals candidate genes involved in infection. mBio. 2012;3(5):e00259–12. Epub 2012/09/07. pmid:22951933; PubMed Central PMCID: PMC3445971.
      • 128. Deng CH, Plummer KM, Jones DAB, Mesarich CH, Shiller J, Taranto AP, et al. Comparative analysis of the predicted secretomes of Rosaceae scab pathogens Venturia inaequalis and V. pirina reveals expanded effector families and putative determinants of host range. BMC Genomics. 2017;18(1):339. Epub 2017/05/04. pmid:28464870; PubMed Central PMCID: PMC5412055.
      • 129. Klosterman SJ, Subbarao KV, Kang S, Veronese P, Gold SE, Thomma BP, et al. Comparative genomics yields insights into niche adaptation of plant vascular wilt pathogens. PLoS Pathog. 2011;7(7):e1002137. Epub 2011/08/11. pmid:21829347; PubMed Central PMCID: PMC3145793.
      • 130. Gazis R, Kuo A, Riley R, LaButti K, Lipzen A, Lin J, et al. The genome of Xylona heveae provides a window into fungal endophytism. Fungal Biol. 2016;120(1):26–42. Epub 2015/12/24. pmid:26693682.
      • 131. Grandaubert J, Bhattacharyya A, Stukenbrock EH. RNA-seq-Based Gene Annotation and Comparative Genomics of Four Fungal Grass Pathogens in the Genus Zymoseptoria Identify Novel Orphan Genes and Species-Specific Invasions of Transposable Elements. G3 (Bethesda). 2015;5(7):1323–33. Epub 2015/04/29. pmid:25917918; PubMed Central PMCID: PMC4502367.
      • 132. Stukenbrock EH, Christiansen FB, Hansen TT, Dutheil JY, Schierup MH. Fusion of two divergent fungal individuals led to the recent emergence of a unique widespread pathogen species. Proceedings of the National Academy of Sciences. 2012;109(27):10954. pmid:22711811
      • 133. Altschul SF, Madden TL, Schäffer AA, Zhang J, Zhang Z, Miller W, et al. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res. 1997;25(17):3389–402. Epub 1997/09/01. pmid:9254694; PubMed Central PMCID: PMC146917.
      • 134. Thompson JD, Higgins DG, Gibson TJ. CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic acids research. 1994;22(22):4673–80. Epub 1994/11/11. pmid:7984417; PubMed Central PMCID: PMC308517.
      • 135. Stamatakis A. RAxML version 8: a tool for phylogenetic analysis and post-analysis of large phylogenies. Bioinformatics. 2014;30(9):1312–3. Epub 2014/01/24. pmid:24451623; PubMed Central PMCID: PMC3998144.
      • 136. R Core Team. R: A language and environment for statistical computing. R Foundation for Statistical Computing. 2019.
        • 137. Kucheryavskiy S. mdatools: Multivariate Data Analysis for Chemometrics. 2019.
        • 138. Warnes GR, Bolker B, Bonebakker L, Gentleman R, Huber W, Liaw A, et al. gplots: Various R Programming Tools for Plotting Data. 2019.
        • 139. Wickham H, Hester J, Francois R. readr: Read Rectangular Text Data. 2018.
        • 140. Fox J, Weisberg S. An {R} Companion to Applied Regression. Sage. 2019.
          • 141. Murtagh F, Legendre P. Ward’s Hierarchical Agglomerative Clustering Method: Which Algorithms Implement Ward’s Criterion? Journal of Classification. 2014;31(3):274–95.
          • 142. Rice P, Longden I, Bleasby A. EMBOSS: the European Molecular Biology Open Software Suite. Trends Genet. 2000;16(6):276–7. Epub 2000/05/29. pmid:10827456.
          • 143. Bailey TL, Boden M, Buske FA, Frith M, Grant CE, Clementi L, et al. MEME SUITE: tools for motif discovery and searching. Nucleic acids research. 2009;37(Web Server issue):W202–8. pmid:19458158; PubMed Central PMCID: PMC2703892.
          • 144. Kautsar SA, Blin K, Shaw S, Navarro-Muñoz JC, Terlouw BR, van der Hooft JJJ, et al. MIBiG 2.0: a repository for biosynthetic gene clusters of known function. Nucleic Acids Research. 2019;48(D1):D454–D8. pmid:31612915
          • 145. Heneghan MN, Yakasai AA, Williams K, Kadir KA, Wasil Z, Bakeer W, et al. The programming role of trans-acting enoyl reductases during the biosynthesis of highly reduced fungal polyketides. Chemical Science. 2011;2(5):972–9.
          • 146. Ehrlich KC, Chang PK, Yu J, Cotty PJ. Aflatoxin biosynthesis cluster gene cypA is required for G aflatoxin formation. Appl Environ Microbiol. 2004;70(11):6518–24. Epub 2004/11/06. pmid:15528514; PubMed Central PMCID: PMC525170.
          • 147. Bhatnagar D, Cary JW, Ehrlich K, Yu J, Cleveland TE. Understanding the genetics of regulation of aflatoxin production and Aspergillus flavus development. Mycopathologia. 2006;162(3):155–66. Epub 2006/09/01. pmid:16944283.
          • 148. Porquier A, Morgant G, Moraga J, Dalmais B, Luyten I, Simon A, et al. The botrydial biosynthetic gene cluster of Botrytis cinerea displays a bipartite genomic structure and is positively regulated by the putative Zn(II)2Cys6 transcription factor BcBot6. Fungal Genet Biol. 2016;96:33–46. Epub 2016/10/22. pmid:27721016.
          • 149. Pinedo C, Wang CM, Pradier JM, Dalmais B, Choquer M, Le Pecheur P, et al. Sesquiterpene synthase from the botrydial biosynthetic gene cluster of the phytopathogen Botrytis cinerea. ACS Chem Biol. 2008;3(12):791–801. Epub 2008/11/28. pmid:19035644; PubMed Central PMCID: PMC2707148.
          • 150. Hamed RB, Gomez-Castellanos JR, Henry L, Ducho C, McDonough MA, Schofield CJ. The enzymes of β-lactam biosynthesis. Natural product reports. 2013;30(1):21–107. Epub 2012/11/09. pmid:23135477.
          • 151. Abe Y, Suzuki T, Ono C, Iwamoto K, Hosobuchi M, Yoshikawa H. Molecular cloning and characterization of an ML-236B (compactin) biosynthetic gene cluster in Penicillium citrinum. Mol Genet Genomics. 2002;267(5):636–46. Epub 2002/08/13. pmid:12172803.
          • 152. Abe Y, Ono C, Hosobuchi M, Yoshikawa H. Functional analysis of mlcR, a regulatory gene for ML-236B (compactin) biosynthesis in Penicillium citrinum. Mol Genet Genomics. 2002;268(3):352–61. Epub 2002/11/19. pmid:12436257.
          • 153. Hoffmann K, Schneider-Scherzer E, Kleinkauf H, Zocher R. Purification and Characterization of Eucaryotic Alanine Racemase Acting as Key Enzyme in Cyclosporin Biosynthesis. Biological Chemistry. 1994;269(17):12710–4. pmid:8175682
          • 154. Yang X, Feng P, Yin Y, Bushley K, Spatafora JW, Wang C. Cyclosporine Biosynthesis in Tolypocladium inflatum Benefits Fungal Adaptation to the Environment. Molecular Biology and Physiology. 2018;9(5). pmid:30279281
          • 155. Weber G, Leitner E. Disruption of the cyclosporin synthetase gene of Tolypocladium niveum. Current Genetics. 1994;26:461–7. pmid:7874740
          • 156. Wang B, Kang Q, Lu Y, Bai L, Wang C. Unveiling the biosynthetic puzzle of destruxins in Metarhizium species. Proc Natl Acad Sci U S A. 2012;109(4):1287–92. Epub 2012/01/11. pmid:22232661; PubMed Central PMCID: PMC3268274.
          • 157. Lin H-C, Chooi Y-H, Dhingra S, Xu W, Calvo AM, Tang Y. The Fumagillin Biosynthetic Gene Cluster in Aspergillus fumigatus Encodes a Cryptic Terpene Cyclase Involved in the Formation of β-trans-Bergamotene. Journal of the American Chemical Society. 2013;135(12):4616–9. pmid:23488861
          • 158. Kato N, Suzuki H, Takagi H, Uramoto M, Takahashi S, Osada H. Gene disruption and biochemical characterization of verruculogen synthase of Aspergillus fumigatus. Chembiochem. 2011;12(5):711–4. Epub 2011/03/16. pmid:21404415.
          • 159. Kato N, Suzuki H, Takagi H, Asami Y, Kakeya H, Uramoto M, et al. Identification of cytochrome P450s required for fumitremorgin biosynthesis in Aspergillus fumigatus. Chembiochem. 2009;10(5):920–8. Epub 2009/02/20. pmid:19226505.
          • 160. Maiya S, Grundmann A, Li SM, Turner G. Improved tryprostatin B production by heterologous gene expression in Aspergillus nidulans. Fungal Genet Biol. 2009;46(5):436–40. Epub 2009/04/18. pmid:19373974.
          • 161. Kato N, Suzuki H, Okumura H, Takahashi S, Osada H. A point mutation in ftmD blocks the fumitremorgin biosynthetic pathway in Aspergillus fumigatus strain Af293. Biosci Biotechnol Biochem. 2013;77(5):1061–7. Epub 2013/05/08. pmid:23649274.
          • 162. Proctor RH, Busman M, Seo JA, Lee YW, Plattner RD. A fumonisin biosynthetic gene cluster in Fusarium oxysporum strain O-1890 and the genetic basis for B versus C fumonisin production. Fungal genetics and biology: FG & B. 2008;45(6):1016–26. Epub 2008/04/01. pmid:18375156.
          • 163. Zaleta-Rivera K, Xu C, Yu F, Butchko RA, Proctor RH, Hidalgo-Lara ME, et al. A Bidomain Nonribosomal Peptide Synthetase Encoded by FUM14 Catalyzes the Formation of Tricarballylic Esters in the Biosynthesis of Fumonisins. Biochemistry 2006;45:2561–9. pmid:16489749
          • 164. Butchko RA, Plattner RD, Proctor RH. Deletion Analysis of FUM Genes Involved in Tricarballylic Ester Formation during Fumonisin Biosynthesis. J Agric Food Chem. 2006;54:9398−404. pmid:17147424
          • 165. Butchko RA, Plattner RD, Proctor RH. FUM13 Encodes a Short Chain Dehydrogenase/Reductase Required for C-3 Carbonyl Reduction during Fumonisin Biosynthesis in Gibberella moniliformis. J Agric Food Chem 2003;51:3000−6. pmid:12720383
          • 166. Butchko RA, Plattner RD, Proctor RH. FUM9 is required for C-5 hydroxylation of fumonisins and complements the meitotically defined Fum3 locus in Gibberella moniliformis. Appl Environ Microbiol. 2003;69(11):6935–7. Epub 2003/11/07. pmid:14602658; PubMed Central PMCID: PMC262316.
          • 167. Brown DW, Butchko RA, Busman M, Proctor RH. The Fusarium verticillioides FUM gene cluster encodes a Zn(II)2Cys6 protein that affects FUM gene expression and fumonisin production. Eukaryot Cell. 2007;6(7):1210–8. Epub 2007/05/08. pmid:17483290; PubMed Central PMCID: PMC1951116.
          • 168. Du L, Zhu X, Gerber R, Huffman J, Lou L, Jorgenson J, et al. Biosynthesis of sphinganine-analog mycotoxins. J Ind Microbiol Biotechnol. 2008;35(6):455–64. Epub 2008/01/25. pmid:18214562.
          • 169. Proctor RH, Desjardins AE, Plattner RD, Hohn TM. A Polyketide Synthase Gene Required for Biosynthesis of Fumonisin Mycotoxins in Gibberella fujikuroi Mating Population A. Fungal Genetics and Biology 1999;27:100–12. pmid:10413619
          • 170. Lia Y, Lou L, Cerny RL, Butchko RA, Proctor RH, Shen Y, et al. Tricarballylic ester formation during biosynthesis of fumonisin mycotoxins in Fusarium verticillioides. Mycology. 2013;4(4):179–86. Epub 2014/03/04. pmid:24587959; PubMed Central PMCID: PMC3933019.
          • 171. Studt L, Janevska S, Niehaus EM, Burkhardt I, Arndt B, Sieber CM, et al. Two separate key enzymes and two pathway-specific transcription factors are involved in fusaric acid biosynthesis in Fusarium fujikuroi. Environ Microbiol. 2016;18(3):936–56. Epub 2015/12/15. pmid:26662839.
          • 172. Lin X, Yuan S, Chen S, Chen B, Xu H, Liu L, et al. Heterologous Expression of Ilicicolin H Biosynthetic Gene Cluster and Production of a New Potent Antifungal Reagent, Ilicicolin J. Molecules. 2019;24(12). Epub 2019/06/21. pmid:31216742; PubMed Central PMCID: PMC6631495.
          • 173. Cary JW, Uka V, Han Z, Buyst D, Harris-Coward PY, Ehrlich KC, et al. An Aspergillus flavus secondary metabolic gene cluster containing a hybrid PKS-NRPS is necessary for synthesis of the 2-pyridones, leporins. Fungal Genet Biol. 2015;81:88–97. Epub 2015/06/09. pmid:26051490.
          • 174. Manzoni M, Rollini M. Biosynthesis and biotechnological production of statins by filamentous fungi and application of these cholesterol-lowering drugs. Appl Microbiol Biotechnol. 2002;58(5):555–64. Epub 2002/04/17. pmid:11956737.
          • 175. Zhang W, Cao S, Qiu L, Qi F, Li Z, Yang Y, et al. Functional characterization of MpaG’, the O-methyltransferase involved in the biosynthesis of mycophenolic acid. Chembiochem. 2015;16(4):565–9. Epub 2015/01/30. pmid:25630520.
          • 176. Zhang W, Du L, Qu Z, Zhang X, Li F, Li Z, et al. Compartmentalized biosynthesis of mycophenolic acid. Proc Natl Acad Sci U S A. 2019;116(27):13305–10. Epub 2019/06/19. pmid:31209052; PubMed Central PMCID: PMC6613074.
          • 177. Hansen BG, Genee HJ, Kaas CS, Nielsen JB, Regueira TB, Mortensen UH, et al. A new class of IMP dehydrogenase with a role in self-resistance of mycophenolic acid producing fungi. BMC Microbiol. 2011;11:202. Epub 2011/09/20. pmid:21923907; PubMed Central PMCID: PMC3184278.
          • 178. Hansen BG, Salomonsen B, Nielsen MT, Nielsen JB, Hansen NB, Nielsen KF, et al. Versatile enzyme expression and characterization system for Aspergillus nidulans, with the Penicillium brevicompactum polyketide synthase gene from the mycophenolic acid gene cluster as a test case. Appl Environ Microbiol. 2011;77(9):3044–51. Epub 2011/03/15. pmid:21398493; PubMed Central PMCID: PMC3126399.
          • 179. Regueira TB, Kildegaard KR, Hansen BG, Mortensen UH, Hertweck C, Nielsen J. Molecular basis for mycophenolic acid biosynthesis in Penicillium brevicompactum. Appl Environ Microbiol. 2011;77(9):3035–43. Epub 2011/03/15. pmid:21398490; PubMed Central PMCID: PMC3126426.
          • 180. Hansen BG, Mnich E, Nielsen KF, Nielsen JB, Nielsen MT, Mortensen UH, et al. Involvement of a natural fusion of a cytochrome P450 and a hydrolase in mycophenolic acid biosynthesis. Appl Environ Microbiol. 2012;78(14):4908–13. Epub 2012/05/01. pmid:22544261; PubMed Central PMCID: PMC3416377.
          • 181. Del-Cid A, Gil-Duran C, Vaca I, Rojas-Aedo JF, Garcia-Rico RO, Levican G, et al. Identification and Functional Analysis of the Mycophenolic Acid Gene Cluster of Penicillium roqueforti. PLoS One. 2016;11(1):e0147047. Epub 2016/01/12. pmid:26751579; PubMed Central PMCID: PMC4708987.
          • 182. Gillot G, Jany JL, Dominguez-Santos R, Poirier E, Debaets S, Hidalgo PI, et al. Genetic basis for mycophenolic acid production and strain-dependent production variability in Penicillium roqueforti. Food Microbiol. 2017;62:239–50. Epub 2016/11/28. pmid:27889155.
          • 183. Scott B, Young CA, Saikia S, McMillan LK, Monahan BJ, Koulman A, et al. Deletion and gene expression analyses define the paxilline biosynthetic gene cluster in Penicillium paxilli. Toxins (Basel). 2013;5(8):1422–46. Epub 2013/08/21. pmid:23949005; PubMed Central PMCID: PMC3760044.
          • 184. Fierro F, Garcia-Estrada C, Castillo NI, Rodriguez R, Velasco-Conde T, Martin JF. Transcriptional and bioinformatic analysis of the 56.8 kb DNA region amplified in tandem repeats containing the penicillin gene cluster in Penicillium chrysogenum. Fungal Genet Biol. 2006;43(9):618–29. Epub 2006/05/23. pmid:16713314.
          • 185. Xu X, Liu L, Zhang F, Wang W, Li J, Guo L, et al. Identification of the first diphenyl ether gene cluster for pestheic acid biosynthesis in plant endophyte Pestalotiopsis fici. Chembiochem. 2014;15(2):284–92. Epub 2013/12/05. pmid:24302702.
          • 186. Chen L, Yue Q, Zhang X, Xiang M, Wang C, Li S, et al. Genomics-driven discovery of the pneumocandin biosynthetic gene cluster in the fungus Glarea lozoyensis. BMC Genomics. 2013;14(339). pmid:23688303
          • 187. Chen L, Li Y, Yue Q, Loksztejn A, Yokoyama K, Felix EA, et al. Engineering of New Pneumocandin Side-Chain Analogues from Glarea lozoyensis by Mutasynthesis and Evaluation of Their Antifungal Activity. ACS Chem Biol. 2016;11(10):2724–33. Epub 2016/10/22. pmid:27494047; PubMed Central PMCID: PMC5502478.
          • 188. Chen L, Yue Q, Li Y, Niu X, Xiang M, Wang W, et al. Engineering of Glarea lozoyensis for exclusive production of the pneumocandin B0 precursor of the antifungal drug caspofungin acetate. Appl Environ Microbiol. 2015;81(5):1550–8. Epub 2014/12/21. pmid:25527531; PubMed Central PMCID: PMC4325176.
          • 189. Salo O, Guzman-Chavez F, Ries MI, Lankhorst PP, Bovenberg RAL, Vreeken RJ, et al. Identification of a Polyketide Synthase Involved in Sorbicillin Biosynthesis by Penicillium chrysogenum. Appl Environ Microbiol. 2016;82(13):3971–8. Epub 2016/04/24. pmid:27107123; PubMed Central PMCID: PMC4907180.
          • 190. Guzman-Chavez F, Salo O, Nygard Y, Lankhorst PP, Bovenberg RAL, Driessen AJM. Mechanism and regulation of sorbicillin biosynthesis by Penicillium chrysogenum. Microb Biotechnol. 2017;10(4):958–68. pmid:28618182; PubMed Central PMCID: PMC5481523.
          • 191. Derntl C, Guzman-Chavez F, Mello-de-Sousa TM, Busse HJ, Driessen AJM, Mach RL, et al. In Vivo Study of the Sorbicillinoid Gene Cluster in Trichoderma reesei. Frontiers in microbiology. 2017;8:2037. pmid:29104566; PubMed Central PMCID: PMC5654950.
          • 192. Heneghan MN, Yakasai AA, Halo LM, Song Z, Bailey AM, Simpson TJ, et al. First heterologous reconstruction of a complete functional fungal biosynthetic multigene cluster. Chembiochem. 2010;11(11):1508–12. Epub 2010/06/25. pmid:20575135.
          • 193. Halo LM, Heneghan MN, Yakasai AA, Song Z, Williams K, Bailey AM, et al. Late Stage Oxidations during the Biosynthesis of the 2-Pyridone Tenellin in the Entomopathogenic Fungus Beauveria bassiana. J Am Chem Soc. 2008;130:17988–96. pmid:19067514
          • 194. Zaehle C, Gressler M, Shelest E, Geib E, Hertweck C, Brock M. Terrein biosynthesis in Aspergillus terreus and its impact on phytotoxicity. Chem Biol. 2014;21(6):719–31. Epub 2014/05/13. pmid:24816227.
          • 195. Kakule TB, Zhang S, Zhan J, Schmidt EW. Biosynthesis of the Tetramic Acids Sch210971 and Sch210972. Organic Letters. 2015;17(10):2295–7. pmid:25885659
          • 196. Umemura M, Nagano N, Koike H, Kawano J, Ishii T, Miyamura Y, et al. Characterization of the biosynthetic gene cluster for the ribosomally synthesized cyclic peptide ustiloxin B in Aspergillus flavus. Fungal Genet Biol. 2014;68:23–30. Epub 2014/05/21. pmid:24841822.
          • 197. Lim FY, Won TH, Raffa N, Baccile JA, Wisecaver J, Rokas A, et al. Fungal Isocyanide Synthases and Xanthocillin Biosynthesis in Aspergillus fumigatus. mBio. 2018;9(3). Epub 2018/05/31. pmid:29844112; PubMed Central PMCID: PMC5974471.
          • 198. Chicco D, Jurman G. The advantages of the Matthews correlation coefficient (MCC) over F1 score and accuracy in binary classification evaluation. BMC Genomics. 2020;21(1):6. Epub 2020/01/04. pmid:31898477; PubMed Central PMCID: PMC6941312.
          • 199. Frishman WH, Rapier RC. Lovastatin: an HMG-CoA reductase inhibitor for lowering cholesterol. The Medical clinics of North America. 1989;73(2):437–48. Epub 1989/03/01. pmid:2645482.
          • 200. Hutchinson CR, Kennedy J, Park C, Kendrew S, Auclair K, Vederas J. Aspects of the biosynthesis of non-aromatic fungal polyketides by iterative polyketide synthases. Antonie van Leeuwenhoek. 2000;78(3):287–95. pmid:11386351
          • 201. Gauglitz G. Artificial vs. human intelligence in analytics. Analytical and bioanalytical chemistry. 2019;411(22):5631–2. pmid:31240356
          • 202. Buchfink B, Xie C, Huson DH. Fast and sensitive protein alignment using DIAMOND. Nature methods. 2015;12(1):59–60. pmid:25402007
          • 203. Finn RD, Clements J, Eddy SR. HMMER web server: interactive sequence similarity searching. Nucleic acids research. 2011;39(Web Server issue):W29–W37. Epub 05/18. pmid:21593126.