2.9 | Gene content in the candidate genomic regions
The annotated reference genome of P. fijiensis was used to identify all the genes in all the genomic regions detected using the genome scan approaches and that were putatively involved in the host adaptation. The annotation (Go terms, KOG terms, and presence of a peptide signal) for these genes was retrieved from the GGF3 file of the JGI website (https://genome.jgi.doe.gov/Mycfi2/Mycfi2.home.html). We also tested whether these genes corresponded to in silico defined SSPs (Arango Isaza et al., 2016) or to proteins secreted in vitro and in plantaby comparing isolates with different pathogenicity levels (Escobar-Tovar et al., 2015) or genes expressed during infection in a transcriptome analysis (Noar & Daub, 2016). Protein sequences were subjected to a BLAST (Altschul et al., 1990) search in the pathogen-host interactions database (PHI-base, version 4.9), which currently contains around 6 000 genes proven to affect the outcome of host-pathogen interactions (Urban et al., 2015). For each BLAST search, we kept the gene in the PHI-base that had the lowest bitscore (and E value < 1x10-9) and an effect on quantitative pathogenicity (i.e. phenotype “loss of pathogenicity”, “reduced virulence”, “increased virulence”).