Optimized analysis parameters of variant calling for whole genome-based phylogeny of Mycobacteroides abscessus.

Whole-genome sequence (WGS) analysis provides the best resolution for reconstructing bacterial phylogeny. However, the resulting tree could vary according to parameters used in the WGS pipeline, making it difficult to compare results across multiple studies. This study compares effects on phylogenies when applying different parameter stringencies. We used as the study model to optimize parameters strains of Mycobacteroides abscessus serially isolated at various intervals, isolates known to represent persistent infection (PI) cases or re-infection (RI) cases and isolates from different subspecies. Un-optimized parameters with low stringency provided an excessive number of SNPs (823) compared to the optimized setting (3 SNPs) between paired strains isolated 1 day apart from PI cases, discordant tree topology and misclassification of subspecies and of instances of RI. We demonstrated that using high-quality variants provides more accuracy for recognizing serial isolates of the same clone versus different clones and for phylogenetic analysis of M. abscessus. Our approach might be used as a model for analyses requiring phylogenetic reconstruction of other bacteria.