Pangenome gene alignments of 571 Rhizobiaceae

All predicted protein sequence were extracted from the 571Rhizob dataset and clustered into homologous gene families (with possibly multiple copies per genome) using MMSeqs2 release 2 (commit e5d64b2) (Steinegger & Söding 2017).Protein sequences from each gene family were aligned using ClustalOmega (Sievers et al. 2011) and reverse-translated into coding sequence (CDS) alignments using PAL2NAL (Suyama et al. 2006). This resulted in a set of 50,791 final non-ORFan CDS families, including 9,589 homogeneous (derived from identical protein sequences) families.