In this paper we introduce Armadillo v1.1, a novel workflow platform dedicated to designing and conducting phylogenetic studies, including comprehensive simulations. A number of important phylogenetic and general bioinformatics tools have been included in the first software release. As Armadillo is an open-source project, it allows scientists to develop their own modules as well as to integrate existing computer applications. Using our workflow platform, different complex phylogenetic tasks can be modeled and presented in a single workflow without any prior knowledge of programming techniques. The first version of Armadillo was successfully used by professors of bioinformatics at Université du Quebec à Montreal during graduate computational biology courses taught in 2010–11. The program and its source code are freely available at: <http://www.bioinfo.uqam.ca/armadillo>.
Citation: Lord E, Leclercq M, Boc A, Diallo AB, Makarenkov V (2012) Armadillo 1.1: An Original Workflow Platform for Designing and Conducting Phylogenetic Analysis and Simulations. PLoS ONE7(1): e29903. https://doi.org/10.1371/journal.pone.0029903
Editor: Simon Joly, Montreal Botanical Garden, Canada
Received: June 16, 2011; Accepted: December 8, 2011; Published: January 11, 2012
Copyright: © 2012 Lord 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.
Funding: EL and AB are Natural Science and Engineering Research Council of Canada fellows (NSERC). VM and ABD hold NSERC discovery grants. 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.
Bioinformatics is a fast-evolving field that encompasses molecular biology, biochemistry, computer science, mathematics and statistics . It has emerged as a field of its own due to the dramatic increase and complexity of available genomic data . Phylogenetics, which is a subfield of bioinformatics and molecular biology, studies evolutionary relationships between organisms based on their molecular or morphological proximity and presents those relationships through illustrations called phylogenetic trees (or phylogenies) .
The development of a variety of bioinformatics and phylogenetic algorithms has led to the design of numerous computer applications often generating different results when solving the same computational biology problem , . Thus, modeling and performing computational biology simulations can be a very challenging task due to the quantity and diversity of available tools and genomic databanks. Tutorials and examples have been commonly distributed with phylogenetic and bioinformatics applications, whereas online training courses can be found in different web directories, such as Bioinformatics Link Directory  for example. However, “standard practices” for bioinformatics and phylogenetic analysis have not been strictly defined (apart from a number of specific fields , ); each analytical step can be carried out using a variety of methods and tools . Moreover, when conducting their experiments and simulations, computational biologists have to cope with programs' limitations and data integration issues . In addition, incorrect outcomes of biological data analysis can arise when accessible, but wrong, tools and models are used .
Meanwhile, the increasing use of genomic and phylogenetic data fuels the need for pipeline managing software. For instance, Cicarelli et al.  developed an automatable procedure for reconstructing the tree of life with branch lengths comparable across the three species domains which can be represented as a typical bioinformatics workflow encompassing the following major tasks (also see Fig. 1. in ): Selection and preparation of marker gene families, generation of concatenated multiple sequence alignments (supermatrix approach), systematic detection and removal of horizontal gene transfers, phylogenetic tree reconstruction and evaluation. Philippe et al.  warn, however, that automated phylogenetic analysis has its own pitfalls and caveats. Thus, all automatically generated data must be “manually” verified and corrected, if necessary, before proceeding to their further treatment or interpretation.
Panel (A) presents a standard pipeline using Perl scripting; Panels (B, C, D) show different workflow designs for the sequence search operation provided by Galaxy (B) , Taverna (C)  and the introduced Armadillo workflow platform (D).
A typical bioinformatics task can be described as a workflow pipeline in which resources or data are sequentially processed by a series of dedicated tools . A workflow pipeline usually includes: (1) Data acquisition, (2) data analysis and (3) report and results generation steps . Figure 1 presents an example of a standard task carried out when searching for DNA sequences similar to a given one. Such a task requires a combination of three steps: (1) Input of the given DNA sequences in a particular data format, (2) its conversion to a compatible format according to the tools being used, and finally, (3) execution of an appropriate online or local database query using a sequence matching algorithm such as BLAST . Hence, building solid bioinformatics studies requires the integration of several computational tools, which can include public or commercial computer applications. The following major issues can arise while integrating them into a single pipeline: Communication protocols between applications (e.g., file formats and command line options), computational requirements of embedded algorithms, limited or inexistent electronic access to public or private biological databanks, software running under different operating systems and non-standardized result reports generated by the applications . Life scientists with no prior knowledge in computer science are either limited to the basic usage of existing tools or need to learn a script language, such as Perl (Fig. 1A), or a programming language, such as C or Java , in order to implement their experiments.
Several computational tools have been introduced to help bioinformatics researchers perform complex tasks: (1) Web-based applications, such as Galaxy  (Fig. 1B) or ENSEMBL-European Bioinformatics Institute platform (EBI) , providing users with a multi-processor cluster access and user-friendly interface. In the same context, web-only portals such as Phylogeny.fr , Phylemon  and Bioextract.org  allows the creation of simple execution pipelines intended for phylogenetic inference (e.g., loops and conditional operations are not allowed in those pipelines), along with the AIR  web server, which is intended for conducting phylogenomics analysis using supermatrix methodology; (2) Dedicated programming libraries, BioPerl , BioJava , etc., written in popular script languages in order to facilitate programming tasks; (3) Multi-task applications (i.e., applications merging different programs) or computer packages including a variety of general bioinformatics algorithms, such as MEGA , Geneious  and Mesquite . The examples of bioinformatics tasks that can be performed using the three latter packages include multiple sequence alignment, phylogenetic inference, short-read assembly, recombination analysis, reconstruction of ancestral states, simulation of character evolution and detection of deep coalescence. In the same way, the packages PAML , PAUP , PHYLIP  and T-Rex  are specialized multifunction software focusing on phylogenetic analysis.
While the first and third types of tools rely on the manipulation of diverse data formats and manual execution of different applications to answer real-life biological questions, the second type of tools requires deep knowledge of programming languages. An emerging approach in the Life Science community is the development of workflow platforms (Fig. 1B–1D)  which can be used for designing and running comprehensive simulation studies as well as for educational purposes.
Workflow platforms consist of connected processes or applications. They were originally implemented and used in financial and business environments , . Workflows have been also actively employed to chain together specialized programs in order to create dataflow for software development, including scientific applications . They provide a suitable environment for defining and managing the coordination and allocation of process activities . Workflow platforms expand from simple execution managers to complex systems often allowing for conditional dataflow scheduling and task distribution , . Such platforms also simplify the automation of error-prone tasks, data collection including refactoring and organization of input/output, data processing and results visualization , . The examples of workflow platforms dedicated to bioinformatics include the web-based platform Galaxy  (Fig. 1B) and desktop-based platform Taverna  (Fig. 1C). Both of them contain a specific workflow language and particular computational model supporting automation. While Galaxy is based on private server architecture, Taverna relies on well developed web-services.
Previous studies showed that workflow platforms can be also useful in finding solutions, answering questions and facilitating “just-in-time” learning and demonstration . Several workflow platforms have been dedicated to education , –. They provide teachers with different web-based (e-learning, tele-learning) or desktop-based protocols helping them create appropriate educational content , , while keeping record of students' results, failures and attempts . Most of educational workflows intend either to enhance cognition of students or verify the quality of the learning process . To the best of our knowledge, no workflow platforms dedicated to bioinformatics, or phylogenetic, education have been proposed. Thus, the development of a workflow application encapsulating and managing the most popular bioinformatics algorithms and data formats would be a very important task for this vital Life Science field. Such a platform would permit students to ignore the “cooking” programming and to focus directly on actual goals of their projects. It would ideally satisfy to a WYSIWYG (What You See Is What You Get) principle, or more precisely in our case, to a WYPIWYG (What You Pipe Is What You Get) principle .
In this article we describe Armadillo, an original workflow platform dedicated to modeling complex phylogenetic as well as certain general bioinformatics experiments. Armadillo proceeds by a local execution of popular phylogenetic applications. It allows users to quickly design large-scale computational biology experiments, facilitating data format management, electronic transactions with biological databanks and manual conversion of inputs and outputs of several major bioinformatics applications. Our platform proposes an elegant graphical user interface (GUI), conventionally presenting the available applications as single boxes and computational pipelines as box interconnections. Interconnections represent the data acquisition events and create a flow involving different existing applications (see Fig. 1D). In the following sections we will describe how Armadillo can be used to carry out phylogenetic analysis and facilitate phylogenetic and general bioinformatics learning.
General description of Armadillo
The Armadillo workflow platform (version 1.1) has been developed using the processing library for GUI operations (Fig. 2). Similarly to the other existing bioinformatics platforms, such as Taverna  or Galaxy , the components representing either datasets or bioinformatics applications can be linked together to create a dataflow by performing drag-and-drop box-connection operations (Fig. 2A and 2B). The Armadillo platform integrates a sequence viewer and provides access to phylogenetic tree interference and manipulation applications (Fig. 2C and 2D). The configuration of each application is performed via a custom dialog box (Fig. 2E) which facilitates the use of the most commonly used features. Armadillo does not require an Internet access for most of its operations. Developed in Java, it runs under the Windows OS and Mac OS X (including Leopard and Lion) operating systems and comprises several popular Windows and Apple-Macintosh compiled applications. The minimum system requirements are stated on the home page of the Armadillo website (for more details, see: http://www.bioinfo.uqam.ca/armadillo). The installation of Armadillo can be performed in a single step which includes the set up of many popular phylogenetic programs (see Table 1). Certainly, the user has to cope with particular memory and parameters limitations imposed by the applications included in the Armadillo platform as well as with the RAM overflow that can be caused by executing those applications on large datasets. By using the Preference→Advanced option, the user can specify the maximum amount of RAM memory that can be used by Java.
Panel (A) presents the available tools. All these tools can be used as drag-and-drop components in the workflow. Panel (B) presents the main view of the workflow design. Panel (C) presents a picture of a phylogenetic tree (i.e., phylogeny or evolutionary tree) displayed using the PhyloWidget application . Panel (D) shows an integrated sequence viewer. Panel (E) shows an example of a custom component view allowing an easy configuration of the user pipeline.
As highlighted by Stevens et al. , life cycle of in silico experiments consists of different key points that need to be addressed, including: Performing experimental design, carrying out executions, interpreting the obtained results and presenting or publishing related articles or research reports. The Armadillo platform has been built in order to ease many of such in silico tasks. Mention that all application inputs taken from application outputs are written into a single “project” storage file. Storage files are implemented by means of a Structured Query Language (SQL) compatible database, built using the Xerial Java implementation of SQLite . The project file database can also be displayed on a website to allow users to share their results or compare different workflow strategies. Not supporting distributed databases can seem odd in our cloud computer epoch , but the main goal of this project was to ease the phylogenetic workflow design without having to create users, groups or ensure a secure access to databanks. However, we consider all these options as possible future extensions of our workflow platform. The single file system used in Armadillo also facilitates the organization of course material allowing for a quick creation of exercises, datasets and working diagrams. Text and HTML files can be directly added to any Armadillo workflow. The platform also supports annotation of individual data components, and thus permits teachers, students and researchers to comment the results of their analyses. Finally, the Armadillo architecture can be also used during an SQL language course since all SQL queries are displayed and executed directly within the platform.
Applications included in Armadillo 1.1
The first version of the Armadillo platform supports various sequence data formats via the inclusion of the ReadSeq application developed by Gilbert . For example, the Newick format is used as a default format to represent phylogenetic trees. The tree visualization can be generated using either the PhyloWidget viewer  or the Archaeopteryx application  supporting the XML, Tree-of-Life and Nexus file formats. Several popular tools for solving fundamental phylogenetic problems, such as reconstruction of species evolution from molecular sequences (i.e., phylogenetic inferring), multiple sequence alignment algorithms and database queries using BLAST constitute the core applications included in Armadillo 1.1 (see Table 1 for the complete list of available applications). For instance, the PAML  and PROTML (from the PHYLIP package ) programs provided with Armadillo can deal with different codon models. The Gblocks application ,  can be used to improve the quality of multiple sequence alignments through removing divergent and ambiguously aligned blocks. Custom functions can be also added to the workflow since the runtime compilation and execution of Java programming source code are integrated in our platform. This feature is available via the Tools→Your Program→Custom program option. A description of the basic functions of each application included in Armadillo is accessible by clicking the application information button; this button is represented by the “?” symbol.
The results generated by any application embedded in an Armadillo computational pipeline can be accessed and verified as well. Our platform proceeds by the validation of results of each application during the workflow execution. Once the computation is over, a complete report is generated to present the detailed descriptions of different steps as well as the corresponding results (Fig. 3). Such a report includes all obtained results, application states, error encountered (if any) and application outputs obtained at all intermediate steps of the analysis. Reports are presented through the hyperlinks to the associated HTML files. An example of a complex task used in our simulations is presented in Figure 3. Here the multiple sequence alignment applications (i.e., Muscle  and ProbCons ) provide the input sequence alignments for the phylogenetic tree reconstruction algorithms (i.e., PhyML  and NJ ), which then can be followed by several applications for validating and comparing the obtained tree topologies.
Panel (A) presents available comments and support files (available in the text and HTML formats). Panel (B) presents the beginning of the workflow and the if control used to select between different alternatives in the dataflow. Panel (C) shows how different multiple sequence alignment applications can be modeled. Panel (D) illustrates the use of different colors to annotate different parts of the workflow in order to facilitate the learning process. Panel (E) presents an example of a phylogenetic pipeline. Panel (F) displays an example of obtained results (i.e., results report).
It is worth noting that scientific workflows are usually data-intensive . They can take advantage of the Web 2.0 features either through datasets acquisition, or through distributed computational services. This allows users to free local workstation computational resources for further analyses. We also included to the first version of Armadillo the access to two major data source and web service providers: GenBank maintained by the National Center for Biotechnology Information (NCBI) and Ensembl maintained by European Bioinformatics Institute (EBI) and the Wellcome Trust Sanger Institute (WTSI) (see Table 1).
In Table 2, we compare the main features of the introduced Armadillo platform with those available in the other popular bioinformatics workflow software, including Ergatis , Galaxy , Kepler , LONI  and Taverna . They concern workflow design, experimental data organization and possibility of adding new applications.
Case study: Using Armadillo to infer phylogenetic trees
Phylogenetic tree inference is an important step of many bioinformatics projects such as those dedicated to detection of horizontal gene transfers –. The case study we present here highlights different steps which are necessary for building a phylogeny of the adiponectin protein using the Armadillo platform. Adiponectin, also denoted as Acrp30, apM1, GBP28 or ADIPOQ, due to its discovery by four different research groups, is a 244-amino acid protein primarily secreted by the white adipose tissue. This protein is known for its pleiotropic effects and is involved in obesity-related disorders: Type 2 diabetes mellitus, metabolic syndrome and atherosclerosis. Recently, it has been suggested that adiponectin can also be cardioprotective and has anticancer properties .
Step I: Creating an adiponectin protein dataset.
Inferring a phylogenetic tree is a three-fold process including: (1) creation of a sequence dataset, (2) alignment of the selected sequence data, and (3) inference of a phylogenetic tree from the obtained multiple sequence alignment using a relevant evolutionary model to represent evolutionary history of the considered organisms . The first step here involves searching for the amino acid sequences to be aligned. Starting with a new empty project (File→New project), one can use the main application menu: Manager→Sequences to open Sequence Manager (Fig. 4A). In the sequence manager dialog box, the option: File→Import from Internet should be used. This will open a new dialog box named Search Internet which allows access to data from three major Internet sequence databanks: HUGO (Human Gene Nomenclature Committee), GenBank (supported by NCBI) and Ensembl (see Table 1). For the presentation purpose, we will choose in the database dropdown menu the NCBI Protein database option with the keyword Adiponectin. Once the database search is over, one can filter the search results by retaining for further analysis only the adiponectin sequences and eliminating the derived peptides and adiponectin receptors (i.e., AdipoR1, AdipoR2). To proceed, one has to enter in the Filter your sequences field the range “240–244” and order the sequences by Description. Following the filtering operation, the selected sequences can be downloaded from GenBank using the option Import selected sequences located at the bottom of the dialog box. The sequence manager can be closed afterwards.
Step A: Search dialog box allowing for direct access to different Internet databases. Step B: Creating and interconnecting individual components by means of drag-and-drop operations. Muscle and ProbCons multiple sequence alignment applications are presented here. Step C: Representing the aligned sequences using an internal sequence viewer. Step D: Configuring the options of the PhyML and ProtDist applications prior to phylogenetic inference. Step E: Visualizing the resulting PhyML phylogenetic tree using the Archaeopteryx tree viewer. Panel (F): Displaying the complete computational workflow after a sequential execution of the first (multiple sequence alignment algorithms) and second (phylogenetic tree inference algorithms) workflow parts.
Step II. Generating multiple protein sequence alignments.
Generation of a correct alignment of the selected sequences is the second key step of the tree building protocol , . To align the downloaded protein sequences, one can go to the left workflow toolbox (Fig. 4B), select the newly added MultipleSequences object in the Tool→Database panel and drag and drop it to the workflow creation area (WFCA). Two multiple sequence alignment applications were used in this example to treat the adiponectin protein dataset. First of them is the Muscle program , which usually shows a greater accuracy than the popular ClustalW algorithm  when aligning protein sequences, and the second one, the ProbCons program  based on probabilistic modeling. To use those programs, one need to incorporate them into the workflow by going into the Toolbox→Tools (Fig. 2A) panel and expanding the Tools→Alignment subtree (Fig. 4B). Once the Muscle and ProbCons applications are located, they can be dragged and dropped onto WFCA (Fig. 2B) and connected to the previously added MultipleSequences object. One can now execute the workflow using either the Run button located at the top-right corner of WFCA (Fig. 2B) or the Execution→Run option from the main application menu. Once all execution steps are completed, the progress bar at the top of WFCA will advance to the 100% mark (Fig. 2B). One can visualize the obtained results (Fig. 4C) by either expanding the program output in WFCA and then double-clicking on the resulting alignment or by going into the workflow Toolbox→Current Workflow panel, then expanding the Muscle and ProbCons programs subtree, right-clicking on the alignment of choice, and finally selecting the Details or View Graphic option in the contextual pop-up menu.
Step III. Inferring the adiponectin phylogenetic trees.
Once the user is satisfied with one of the obtained sequence alignment, he/she can start inferring phylogenies. Here we present two alternative solutions: The first tree will be obtained using a fast a very popular maximum likelihood method implemented in the PhyML application , while the second tree will be generated by the Neighbor-Joining method  available in the PHYLIP package (i.e., the Neighbor application ). To perform the tree inferring, one can go to the Toolbox→Tools panel, select the Tree→PhyML option, drag and drop the application onto WFCA and then connect it to the newly created alignment. In the same way, one can go to the Tree→PHYLIP subtree, drag and drop the Seqboot, Protdist and Neighbor applications onto WFCA and thus create a second tree inferring pipeline by connecting these applications as shown in Figure 4F. To customize the program execution parameters, one can double-click on the corresponding box of WFCA. For example, in the case of the PhyML program and adiponectin dataset, one could select the following parameters in the default option area: aa (protein) - for the dataset, JTT model - for the substitution, and 100 - for the number of bootstrap replicates (see Fig. 4D). Alternatively, using a PHYLIP package application one could also perform the Model Testing→ProtTest  (see Fig. 3) test to identify the most appropriate evolutionary model characterizing the considered multiple sequence alignment. Once all execution steps are completed, the resulting phylogenetic trees (in the Newick format) can be viewed by double-clicking on the generated output file. On the other hand, a graphical representation of the obtained phylogenetic trees can be generated by right-clicking on the tree output file and selecting the option View Tree in Archaeopteryx or View Tree in PhyloWidget in the contextual pop-up menu (see Fig. 4E). Obviously, the whole computational pipeline as well as the obtained intermediate and final results can be saved and then easily modified to carry out further analyses.
Results and Discussion
In this article we described Armadillo (version 1.1), an original workflow platform dedicated to designing and performing phylogenetic analysis and simulations. The new platform implements an intuitive approach for task automation and simulation design. The current version of Armadillo allows the creation of repetitive workflows, while ensuring the compatibility of numerous phylogenetic and certain general bioinformatics applications distributed with the platform. For instance, Armadillo can be used to perform multiple gene analyses (for example, by means of the for loop allowing for processing multiple gene files). Armadillo was first used to design and carry out a simulation study comparing the most popular methods for inferring horizontal gene transfer events (i.e., HGT-Detection , RIATA_HGT  and LatTrans  algorithms were compared in a 2010 Systematic Biology article ). The new platform was then successively employed as a support tool during the 2010–11 academic year by professors of bioinformatics at Université du Québec à Montréal in the framework of several graduate computational biology courses. Armadillo has been also recently used to perform a complex microRNA classification study conducted by the bioinformaticians of Université du Québec à Montréal as well as a HIV selective pressure study carried out by the researchers of Sainte-Justine Hospital of Montreal. Our software provides a number of interesting features unavailable in the Taverna  and Galaxy  workflow platforms (see Table 2) and includes the tools for solution sharing, genomic databank transactions and program integration. The current version of Armadillo has been released under the Open Source GNU General Public License. The source code, several tutorials and application examples are provided on the Armadillo website: http://www.bioinfo.uqam.ca/armadillo. It is worth noting that we have also implemented a Wiki-style user guide in order to enhance the software learning capacities. Such an interactive manual allows users to suggest the addition of new options and applications as well. Mention that Armadillo can be easily extended by other developers. This extension can concern either the workflow platform itself or the integrated data types and included bioinformatics tools. In the future, we plan to add to Armadillo different multimedia files to increase its learning capacity. We also envisage including in the new platform a data management application and providing users with a data storage facility. All these improvements of Armadillo will aid teachers to create specific educational toolkits for Life Science students. They will also allow Armadillo to become a tool of choice for conducting complex phylogenetic as well as broad-spectrum bioinformatics simulations for many molecular and evolutionary biologists, bioinformaticians and biostatisticians.
The authors are thankful to the editor Simon Joly and to two anonymous reviewers for their helpful comments and suggestions.
Conceived and designed the experiments: EL ML AB ABD VM. Performed the experiments: EL ML AB ABD VM. Analyzed the data: EL VM. Wrote the paper: EL ABD VM.
- 1. Oinn G, Greenwood T, Addis M, Alpdemir MN, Ferris M, et al. (2006) Taverna: Lessons in creating a workflow environment for the life sciences. CCPE 18: 1067–1100.G. OinnT. GreenwoodM. AddisMN AlpdemirM. Ferris2006Taverna: Lessons in creating a workflow environment for the life sciences.CCPE1810671100
- 2. Hoon S, Ratnapu KK, Chia JM, Kumarasamy B, Juguang X, et al. (2003) Biopipe: a flexible framework for protocol-based bioinformatics analysis. Genome Res 13: 1904–1915.S. HoonKK RatnapuJM ChiaB. KumarasamyX. Juguang2003Biopipe: a flexible framework for protocol-based bioinformatics analysis.Genome Res1319041915
- 3. Felsenstein J (2004) Inferring phylogenies. Sinauer Associates, Sunderland, Mass. J. Felsenstein2004Inferring phylogeniesSinauer Associates, Sunderland, Mass
- 4. Stevens R, Zhao J, Goble C (2007) Using provenance to manage knowledge of In Silico experiments. Brief Bioinform 8: 183–194.R. StevensJ. ZhaoC. Goble2007Using provenance to manage knowledge of In Silico experiments.Brief Bioinform8183194
- 5. Brazas MD, Yamada JT, Ouellette BF (2010) Providing web servers and training in Bioinformatics: 2010 update on the Bioinformatics Links Directory. Nucleic Acids Res 38: W3–W6.MD BrazasJT YamadaBF Ouellette2010Providing web servers and training in Bioinformatics: 2010 update on the Bioinformatics Links Directory.Nucleic Acids Res38W3W6
- 6. Yang Z (2007) PAML 4: phylogenetic analysis by maximum likelihood. Mol Biol Evol 24: 1586–1591.Z. Yang2007PAML 4: phylogenetic analysis by maximum likelihood.Mol Biol Evol2415861591
- 7. Swofford DL (2002) PAUP*: Phylogenetic analysis using parsimony (*and other methods). Version 4.0. Sinauer Associates, Sunderland, MA. DL Swofford2002PAUP*: Phylogenetic analysis using parsimony (*and other methods). Version 4.0Sinauer Associates, Sunderland, MA
- 8. Wong KM, Suchard MA, Huelsenbeck JP (2008) Alignment uncertainty and genomic analysis. Science 319: 473–476.KM WongMA SuchardJP Huelsenbeck2008Alignment uncertainty and genomic analysis.Science319473476
- 9. Ciccarelli FD, Doerks T, von Mering C, Creevey CJ, Snel B, et al. (2006) Toward automatic reconstruction of a highly resolved tree of life. Science 311: 1283–1287.FD CiccarelliT. DoerksC. von MeringCJ CreeveyB. Snel2006Toward automatic reconstruction of a highly resolved tree of life.Science31112831287
- 10. Philippe H, Brinkmann H, Lavrov DV, Littlewood DT, Manuel M, et al. (2011) Resolving difficult phylogenetic questions: why more sequences are not enough. PLoS Biol 9: e1000602.H. PhilippeH. BrinkmannDV LavrovDT LittlewoodM. Manuel2011Resolving difficult phylogenetic questions: why more sequences are not enough.PLoS Biol9e1000602
- 11. Johnson M, Zaretskaya I, Raytselis Y, Merezhuk Y, McGinnis S, et al. (2008) NCBI BLAST: a better web interface. Nucleic Acids Res 36: W5–W9.M. JohnsonI. ZaretskayaY. RaytselisY. MerezhukS. McGinnis2008NCBI BLAST: a better web interface.Nucleic Acids Res36W5W9
- 12. Dudley JT, Butte AJ (2009) A Quick Guide for Developing Effective Bioinformatics Programming Skills. PLoS Comput Biol 5: e1000589.JT DudleyAJ Butte2009A Quick Guide for Developing Effective Bioinformatics Programming Skills.PLoS Comput Biol5e1000589
- 13. Giardine B, Riemer C, Hardison R, Burhans R, Elnitski L, et al. (2005) Galaxy: a platform for interactive large-scale genome analysis. Genome Res 15: 1451–1455.B. GiardineC. RiemerR. HardisonR. BurhansL. Elnitski2005Galaxy: a platform for interactive large-scale genome analysis.Genome Res1514511455
- 14. Goujon M, McWilliam H, Li W, Valentin F, Squizzato S, et al. (2010) A new bioinformatics analysis tools framework at EMBL-EBI. Nucl Acids Res 38: W695–W699.M. GoujonH. McWilliamW. LiF. ValentinS. Squizzato2010A new bioinformatics analysis tools framework at EMBL-EBI.Nucl Acids Res38W695W699
- 15. Dereeper A, Guignon V, Blanc G, Audic S, Buffet S, et al. (2008) Phylogeny.fr: robust phylogenetic analysis for the non-specialist. Nucleic Acids Res 36: W465–W469.A. DereeperV. GuignonG. BlancS. AudicS. Buffet2008Phylogeny.fr: robust phylogenetic analysis for the non-specialist.Nucleic Acids Res36W465W469
- 16. Sánchez R, Serra F, Tárraga J, Medina I, Carbonell J, et al. (2011) Phylemon 2.0: a suite of web-tools for molecular evolution, phylogenetics, phylogenomics and hypotheses testing. Nucleic Acids Res 39: W470–W474.R. SánchezF. SerraJ. TárragaI. MedinaJ. Carbonell2011Phylemon 2.0: a suite of web-tools for molecular evolution, phylogenetics, phylogenomics and hypotheses testing.Nucleic Acids Res39W470W474
- 17. Lushbough CM, Jennewein DM, Brendel VP (2011) The BioExtract Server: a web-based bioinformatic workflow platform. Nucleic Acids Res 39: W528–W532.CM LushboughDM JenneweinVP Brendel2011The BioExtract Server: a web-based bioinformatic workflow platform.Nucleic Acids Res39W528W532
- 18. Kumar S, Skjæveland Å, Orr RJ, Enger P, Ruden T, et al. (2009) AIR: A batch-oriented web program package for construction of supermatrices ready for phylogenomic analyses. BMC Bioinformatics 10: 357.S. KumarÅ. SkjævelandRJ OrrP. EngerT. Ruden2009AIR: A batch-oriented web program package for construction of supermatrices ready for phylogenomic analyses.BMC Bioinformatics10357
- 19. Stajich JE, Block D, Boulez K, Brenner SE, Chervitz SA, et al. (2002) The Bioperl toolkit: Perl modules for the life sciences. Genome Res 12: 1611–1618.JE StajichD. BlockK. BoulezSE BrennerSA Chervitz2002The Bioperl toolkit: Perl modules for the life sciences.Genome Res1216111618
- 20. Holland RC, Down TA, Pocock M, Prlić A, Huen D, et al. (2008) BioJava: an open-source framework for bioinformatics. Bioinformatics 24: 2097–2097.RC HollandTA DownM. PocockA. PrlićD. Huen2008BioJava: an open-source framework for bioinformatics.Bioinformatics2420972097
- 21. Tamura K, Peterson D, Peterson N, Stecher G, Nei M, et al. (2011) MEGA5: molecular evolutionary genetics analysis using maximum likelihood, evolutionary distance, and maximum parsimony methods. Mol Biol Evol 28: 2731–2739.K. TamuraD. PetersonN. PetersonG. StecherM. Nei2011MEGA5: molecular evolutionary genetics analysis using maximum likelihood, evolutionary distance, and maximum parsimony methods.Mol Biol Evol2827312739
- 22. Drummond AJ, Ashton B, Cheung M, Heled J, Kearse M, et al. (2009) Geneious, v. AJ DrummondB. AshtonM. CheungJ. HeledM. Kearse2009Geneious, v.5.5; Available: http://www.geneious.com/. Accessed 2011 Oct 26. 5.5; Available: http://www.geneious.com/. Accessed 2011 Oct 26.
- 23. Maddison WP, Maddison DR (2011) Mesquite: a modular system for evolutionary analysis. WP MaddisonDR Maddison2011Mesquite: a modular system for evolutionary analysis.(Version 2.75) Available: http://mesquiteproject.org. Accessed 2011 Oct 26. (Version 2.75) Available: http://mesquiteproject.org. Accessed 2011 Oct 26.
- 24. Felsenstein J (1989) PHYLIP - Phylogeny Inference Package (Version 3.6). Cladistics 5: 164–166.J. Felsenstein1989PHYLIP - Phylogeny Inference Package (Version 3.6).Cladistics5164166
- 25. Makarenkov V (2001) T-REX: reconstructing and visualizing phylogenetic trees and reticulation networks. Bioinformatics 17: 664–668.V. Makarenkov2001T-REX: reconstructing and visualizing phylogenetic trees and reticulation networks.Bioinformatics17664668
- 26. Romano P (2008) Automation of in-silico data analysis processes through workflow management systems. Brief Bioinform 9: 57–68.P. Romano2008Automation of in-silico data analysis processes through workflow management systems.Brief Bioinform95768
- 27. Oinn T, Li P, Kell DB, Goble C, Goderis A, et al. Berlin DE (2007) Taverna/myGrid: Aligning a Workflow System with the Life Sciences Community. Workflows for e-Science, Part III. Germany: Springer-Verlag. pp. 300–319.T. OinnP. LiDB KellC. GobleA. GoderisDE Berlin2007Taverna/myGrid: Aligning a Workflow System with the Life Sciences Community.Workflows for e-Science, Part IIIGermanySpringer-Verlag300319
- 28. Woollard D, Medvidovic N, Gil Y, Mattmann CA (2008) Scientific software as workflows: From discovery to distribution. IEEE Softw 25: 37–43.D. WoollardN. MedvidovicY. GilCA Mattmann2008Scientific software as workflows: From discovery to distribution.IEEE Softw253743
- 29. Lin J, Ho C, Sadiq W, Orlowska ME (2001) On workflow enabled e-learning services. pp. 349–352.J. LinC. HoW. SadiqME Orlowska2001On workflow enabled e-learning services.349352In: Proceedings of the IEEE International Conference on Advanced Learning Technologies, 6–8 Aug. 2001, Madison, WI.
- 30. Beaulah SA, Correll MA, Munro RE, Sheldon JG (2008) Addressing informatics challenges in Translational Research with workflow technology. Drug Discov Today 13: 771–777.SA BeaulahMA CorrellRE MunroJG Sheldon2008Addressing informatics challenges in Translational Research with workflow technology.Drug Discov Today13771777
- 31. Ma J, Shaw E, Kim J (2010) Computational Workflows for Assessing Student Learning. LNCS 6095: 188–197.J. MaE. ShawJ. Kim2010Computational Workflows for Assessing Student Learning.LNCS6095188197
- 32. Deelman E, Singh G, Su M, Blythe J, Gil Y, et al. (2005) Pegasus: A framework for mapping complex scientific workflows onto distributed systems. Scientific Programming 13: 219–237.E. DeelmanG. SinghM. SuJ. BlytheY. Gil2005Pegasus: A framework for mapping complex scientific workflows onto distributed systems.Scientific Programming13219237
- 33. Gil Y, Ratnakar V, Kim J, González-Calero P, Groth P, et al. (2011) WINGS: Intelligent Workflow-Based Design of Computational Experiments. IEEE Intell Syst 26: 62–72.Y. GilV. RatnakarJ. KimP. González-CaleroP. Groth2011WINGS: Intelligent Workflow-Based Design of Computational Experiments.IEEE Intell Syst266272
- 34. van der Veen J, Jones V, Collis B (2000) Using workflow for projects in higher education. Comput Sci Educ 10: 283–301.J. van der VeenV. JonesB. Collis2000Using workflow for projects in higher education.Comput Sci Educ10283301
- 35. Vouk MA, Bitzer DL, Klevans RL (1999) Workflow and end-user quality of service issues in web-based education. IEEE Trans Know Data Eng 11: 673–687.MA VoukDL BitzerRL Klevans1999Workflow and end-user quality of service issues in web-based education.IEEE Trans Know Data Eng11673687
- 36. Zhang FE, Li K (2009) Application of Workflow Technology in Graduate Education Management. pp. 263–266.FE ZhangK. Li2009Application of Workflow Technology in Graduate Education Management.263266In: IEEE '09. International Conference on E-Learning, E-Business, Enterprise Information Systems, and E-Government, 5–6 Dec. 2009,Hong Kong, China.
- 37. Hipp DR, Kennedy D (2003) SQLite. An Embeddable SQL Database Engineer Software. DR HippD. Kennedy2003SQLite. An Embeddable SQL Database Engineer Software.Available: http://www.sqlite.org. Accessed 2011 Oct 26. Available: http://www.sqlite.org. Accessed 2011 Oct 26.
- 38. Yuan D, Yang Y, Liu X, Chen J (2010) A data placement strategy in scientific cloud workflows. Future Gener Comp Sys 26: 1200–1214.D. YuanY. YangX. LiuJ. Chen2010A data placement strategy in scientific cloud workflows.Future Gener Comp Sys2612001214
- 39. Gilbert DG (2010) Readseq by D.G. Gilbert, (ver. 2.1.30). DG Gilbert2010Readseq by D.G. Gilbert, (ver. 2.1.30).Available: http://iubio.bio.indiana.edu/soft/molbio/readseq/java/. Accessed 2011 Oct 26. Available: http://iubio.bio.indiana.edu/soft/molbio/readseq/java/. Accessed 2011 Oct 26.
- 40. Jordan GE, Piel WH (2008) PhyloWidget: web-based visualizations for the tree of life. Bioinformatics 24: 1641–1642.GE JordanWH Piel2008PhyloWidget: web-based visualizations for the tree of life.Bioinformatics2416411642
- 41. Zmasek CM, Eddy SR (2001) ATV: display and manipulation of annotated phylogenetic trees. Bioinformatics 17: 383–384.CM ZmasekSR Eddy2001ATV: display and manipulation of annotated phylogenetic trees.Bioinformatics17383384
- 42. Talavera G, Castresana J (2007) Improvement of phylogenies after removing divergent and ambiguously aligned blocks from protein sequence alignments. Syst Biol 56: 564–577.G. TalaveraJ. Castresana2007Improvement of phylogenies after removing divergent and ambiguously aligned blocks from protein sequence alignments.Syst Biol56564577
- 43. Castresana J (2000) Selection of conserved blocks from multiple alignments for their use in phylogenetic analysis. Mol Biol Evol 17: 540–552.J. Castresana2000Selection of conserved blocks from multiple alignments for their use in phylogenetic analysis.Mol Biol Evol17540552
- 44. Edgar R (2004) MUSCLE: a multiple sequence alignment method with reduced time and space complexity. BMC Bioinformatics 5: 113.R. Edgar2004MUSCLE: a multiple sequence alignment method with reduced time and space complexity.BMC Bioinformatics5113
- 45. Do CB, Mahabhashyam MS, Brudno M, Batzoglou S (2005) ProbCons: Probabilistic consistency-based multiple sequence alignment. Genome Res 15: 330–340.CB DoMS MahabhashyamM. BrudnoS. Batzoglou2005ProbCons: Probabilistic consistency-based multiple sequence alignment.Genome Res15330340
- 46. Guindon S, Gascuel O (2003) A simple, fast, and accurate algorithm to estimate large phylogenies by maximum likelihood. Syst Biol 52: 696–704.S. GuindonO. Gascuel2003A simple, fast, and accurate algorithm to estimate large phylogenies by maximum likelihood.Syst Biol52696704
- 47. Saitou N, Nei M (1987) The neighbor-joining method: a new method for reconstructing phylogenetic trees. Mol Biol Evol 4: 406–425.N. SaitouM. Nei1987The neighbor-joining method: a new method for reconstructing phylogenetic trees.Mol Biol Evol4406425
- 48. Orvis J, Crabtree J, Galens K, Gussman A, Inman JM, et al. (2010) Ergatis: a web interface and scalable software system for bioinformatics workflows. Bioinformatics 26: 1488–1492.J. OrvisJ. CrabtreeK. GalensA. GussmanJM Inman2010Ergatis: a web interface and scalable software system for bioinformatics workflows.Bioinformatics2614881492
- 49. Altintas I, Berkley C, Jaeger E, Jones M, Ludäscher B, et al. (2004) Kepler: An extensible system for design and execution of scientific workflows. Proceedings of Scientific and Statistical Database Management, 21–23 June 2004, Santorini Island, Greece. IEEE Computer Society. pp. 423–442.I. AltintasC. BerkleyE. JaegerM. JonesB. Ludäscher2004Kepler: An extensible system for design and execution of scientific workflows.Proceedings of Scientific and Statistical Database Management, 21–23 June 2004, Santorini Island, GreeceIEEE Computer Society423442
- 50. Dinov ID, Torri F, Macciardi F, Petrosyan P, Liu Z, et al. (2011) Applications of the pipeline environment for visual informatics and genomics computations. BMC Bioinformatics 12: 304.ID DinovF. TorriF. MacciardiP. PetrosyanZ. Liu2011Applications of the pipeline environment for visual informatics and genomics computations.BMC Bioinformatics12304
- 51. Boc A, Philippe H, Makarenkov V (2010) Inferring and validating horizontal gene transfer events using bipartition dissimilarity. Syst Biol 59: 195–211.A. BocH. PhilippeV. Makarenkov2010Inferring and validating horizontal gene transfer events using bipartition dissimilarity.Syst Biol59195211
- 52. Than C, Ruths D, Nakhleh L (2008) PhyloNet: a software package for analyzing and reconstructing reticulate evolutionary relationships. BMC Bioinformatics 9: 322.C. ThanD. RuthsL. Nakhleh2008PhyloNet: a software package for analyzing and reconstructing reticulate evolutionary relationships.BMC Bioinformatics9322
- 53. Addario-Berry L, Hallett M, Lagergren J (2003) Towards identifying lateral gene transfer events. Pac Symp Biocomput 8: 279–290.L. Addario-BerryM. HallettJ. Lagergren2003Towards identifying lateral gene transfer events.Pac Symp Biocomput8279290
- 54. Brochu-Gaudreau K, Rehfeldt C, Blouin R, Bordignon V, Murphy BD, et al. (2010) Adiponectin action from head to toe. Endocrine 37: 11–32.K. Brochu-GaudreauC. RehfeldtR. BlouinV. BordignonBD Murphy2010Adiponectin action from head to toe.Endocrine371132
- 55. Posada D (2008) jModelTest: phylogenetic model averaging. Mol Biol Evol 25: 1253–1256.D. Posada2008jModelTest: phylogenetic model averaging.Mol Biol Evol2512531256
- 56. Thompson J, Higgins D, Gibson T (1994) CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res 22: 4673–4680.J. ThompsonD. HigginsT. Gibson1994CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice.Nucleic Acids Res2246734680
- 57. Abascal F, Zardoya R, Posada D (2005) ProtTest: selection of best-fit models of protein evolution. Bioinformatics 21: 2104–2105.F. AbascalR. ZardoyaD. Posada2005ProtTest: selection of best-fit models of protein evolution.Bioinformatics2121042105
- 58. Valentin F, Squizzato S, Goujon M, McWilliam H, Paern J, et al. (2010) Fast and efficient searching of biological data resources–using EB-eye. Brief Bioinfom 11: 375–384.F. ValentinS. SquizzatoM. GoujonH. McWilliamJ. Paern2010Fast and efficient searching of biological data resources–using EB-eye.Brief Bioinfom11375384
- 59. Suchard MA, Redelings BD (2006) BAli-Phy: simultaneous Bayesian inference of alignment and phylogeny. Bioinformatics 22: 2047–2048.MA SuchardBD Redelings2006BAli-Phy: simultaneous Bayesian inference of alignment and phylogeny.Bioinformatics2220472048
- 60. Larkin MA, Blackshields G, Brown NP, Chenna R, McGettigan P, et al. (2007) Clustal W and Clustal X version 2.0. Bioinformatics 23: 2947–2948.MA LarkinG. BlackshieldsNP BrownR. ChennaP. McGettigan2007Clustal W and Clustal X version 2.0.Bioinformatics2329472948
- 61. Lassmann T, Sonnhammer EL (2005) Kalign - an accurate and fast multiple sequence alignment algorithm. BMC Bioinformatics 6: 298.T. LassmannEL Sonnhammer2005Kalign - an accurate and fast multiple sequence alignment algorithm.BMC Bioinformatics6298
- 62. Olsen GJ, Matsuda H, Hagstrom R, Overbeek R (1994) fastDNAmL: a tool for construction of phylogenetic trees of DNA sequences using maximum likelihood. Comput Appl Biosci 10: 41–48.GJ OlsenH. MatsudaR. HagstromR. Overbeek1994fastDNAmL: a tool for construction of phylogenetic trees of DNA sequences using maximum likelihood.Comput Appl Biosci104148
- 63. Rivas E, Eddy SR (2008) Probabilistic phylogenetic inference with insertions and deletions. PLoS Comput Biol 4: e1000172.E. RivasSR Eddy2008Probabilistic phylogenetic inference with insertions and deletions.PLoS Comput Biol4e1000172
- 64. Zwickl DJ (2006) Genetic algorithm approaches for the phylogenetic analysis of large biological sequence datasets under the maximum likelihood criterion. DJ Zwickl2006Genetic algorithm approaches for the phylogenetic analysis of large biological sequence datasets under the maximum likelihood criterion.Ph.D. dissertation, The University of Texas at Austin. Ph.D. dissertation, The University of Texas at Austin.
- 65. Ronquist F, Huelsenbeck JP (2003) MrBayes 3: Bayesian phylogenetic inference under mixed models. Bioinformatics 19: 1572–1574.F. RonquistJP Huelsenbeck2003MrBayes 3: Bayesian phylogenetic inference under mixed models.Bioinformatics1915721574
- 66. Chevenet F, Croce O, Hebrard M, Christen R, Berry V (2010) ScripTree: scripting phylogenetic graphics. Bioinformatics 26: 1125–1126.F. ChevenetO. CroceM. HebrardR. ChristenV. Berry2010ScripTree: scripting phylogenetic graphics.Bioinformatics2611251126