Fig 1.
The underlying method for AlbaTraDIS.
The inputs are insert site plots, with a frequency count of the insertions at each base in the genome for a condition and controls and the annotated genome in EMBL format. The abundance of inserts are normalised and the plots split into forward strand, reverse strand and combined strand insertions. Essentiality and differential abundance is assessed using sliding windows or a per gene option. The height of the log fold change plot indicates the log fold change difference in insertions between the conditions and controls. The list of significant genes is compiled using user definable values of corrected p value (q-value), logCPM (Logarithm Count Per Million) and logFC (Logarithm Fold Change).
Fig 2.
A) The top four lines are the insertion sites in controls and under treatment conditions, where red lines are insertions in the forward direction and blue lines are insertions in the reverse direction, with the height corresponding to the number mapped reads identified for this site. The next three lines correspond to the signal identified by AlbaTraDIS using a sliding window of 50 bases and an interval of 25 bases, with the height corresponding to the log fold change between the treatments and controls. The bottom section shows the genes as found in the reference genome, with the forward reading frames of translation. B) The pattern of insertions around a gene that imply transcriptional augmentation, in the forward or reverse complementary direction. The shape of the gene indicates the direction, with the 5’ at the beginning (flat end) and the 3’ prime at the pointed end. Insertions on the forward strand are above the line and insertions on the reverse strand are below the line.
Fig 3.
A) Neighbor joining tree of the presence and absence of genes that have significant differences in the number of insertions compared with the control after exposure to different concentrations of Triclosan. This shows how similar different conditions relate to each other based on their modes of action. The tree was drawn using FigTree [24]. B) Example network of the relatedness of different modes of action where the red nodes are different conditions (such as drug concentrations), and the green nodes are a single gene (this network was drawn using Cytoscape [25]).
Fig 4.
The top 4 panels show the transposon insertion sites, 2 for libraries grown in 0.5 mg/L Triclosan and 2 controls.
The next panel shows the signal identified by AlbaTraDIS with the height corresponding to the log fold change between the treatments and controls. There is an increase in insertions in the promotor area (upstream) in the direction towards the gene, which indicates that up-regulation of fabI in E. coli grown in 0.5 mg/L of Triclosan might be beneficial to survival. This shows that AlbaTraDIS can identify the transposon insertions that alter expression of an essential gene.
Fig 5.
The total running time, including comparative analysis, for varying numbers of test conditions when using a single CPU (Blue) and the total memory usage in GB, including comparative analysis, for varying numbers of conditions when using a single CPU (Grey).