E Unibus Plurum: Genomic Analysis of an Experimentally Evolved Polymorphism in Escherichia coli (original) (raw)

< Back to Article

Figure 4

Expression profile SAM analysis of strains in co-culture reflects many, but not all, regulatory changes observed when strains are grown in monoculture.

Genes found to be altered in expression by visually adjusting δ in the consortium 1-class SAM agree with predictions on monoculture results. Predicted expression levels (shown in the far left column for each heat-map) were calculated as a weighted average of monoculture log2 ratios under the assumption that the contribution of each strain to the total RNA pool is proportional to their relative frequency in the chemostat (i.e. 10% CV101, 20% CV116 and 70% CV103). For each panel, columns with an asterisk (*) are significant using the highly stringent method of adjusting δ (see Materials and Methods for details). (A) Heat-map of genes significant in both the consortium (column labeled “cons”) and monoculture 1-class SAM analyses. (B) Genes whose expression is significant in the consortium 1-class SAM analysis and the monoculture 4-class SAM analysis. (C) Genes whose expression is significant in the monoculture 4-class SAM analysis but not in the consortium 1-class SAM analysis. (D) Genes that are significant in the consortium 1-class SAM analysis but not in either of the monoculture analyses. However, the majority of genes in panel D are significant at the less stringent 0% false discovery rate threshold. † to the right of the gene name indicates the gene is not significant at either threshold in any of the monoculture analyses.

Figure 4

doi: https://doi.org/10.1371/journal.pgen.1000713.g004