Bioinformatic and statistical analysis of RNA-Seq data
Trimmed and demultiplexed reads were mapped back to transcripts using Salmon (version 1.1.0) and quantification was imported into R using the tximport package (version 1.10.1) as previously reported . Differential abundance analysis was performed with the DESeq2 version 1.22.2 package . For each S. meliloti strain, genes differentially expressed (log2 fold change of >|2|; P value <0.01) under at least one condition relative to the control conditions were identified, and all fold change values for these genes were extracted. Core and dispensable genome were computed with Roary following annotation with Prokka and assignment to Clusters of Orthologous Genes (COG) categories, as previously reported . Comparison of expression values of genes in the core genome of strains and nested likelihood ratio tests (LRTs) were performed as indicated in . All genes of S. meliloti strains 1021, AK83, and BL225C were functionally annotated using stand- alone version 2 of eggNOG-mapper with default settings.