Multiple Sequence Alignment of a diverse dataset with 1788 Mycobacterium tuberculosis isolates

Multiple Sequence Alignment of a diverse dataset with 1788 Mycobacterium tuberculosis isolates used for ReporTree benchmarking

The dataset comprises whole-genome sequence data published by Walker et al. 2015. For the multiple sequence analysis, we proceeded as follows:

Reads were downloaded from ENA BioProject PRJNA282721 (accessed on March 16th, 2023) and trimmed using Trimmomatic (Bolger et al., 2014) with INNUca default settings;

Quality-processed reads were individually mapped against the H37Rv reference genome (Genbank accession: NC_000962.3) using Snippy v4.5.1 and SNP-calling was performed on variant sites with the following criteria: a minimum proportion of reads differing from the reference of 70%, a minimum mapping quality of 30 and a minimum coverage for SNP calling of 10;

A full alignment was extracted using Snippy’s core module (snippy-core), with masking of SNPs falling within known M. tuberculosis genomic regions with high GC content, repetitive elements and resistance-associated positions (corresponding to ~8% of the genome), as previously described for surveillance purposes (Macedo et al., 2019);

M. tuberculosis lineages were determined using tb-profiler v4.4.1 (Phelan et al., 2019), with samples from the M. tuberculosis complex other than M. tuberculosis, representing a mix of multiple lineages, or with less than 95% of mapped positions in the reference, being excluded;

A filtered alignment comprising the maximum number of informative sites (88,562 nucleotide sites with at least one mutation in a given sequence) was extracted from the full alignment using the alignment_processing.py v1.1.0 (default settings) of ReporTree, and then used as input for the benchmarking.

In this repository, we provide two alignment files:

Core_MTB_1787_strs.full.aln: this corresponds to the full multiple sequence alignment comprising 1787 samples and the reference (corresponding to the point 4 of the methodology).

MTb_original_align_profile.fasta: this corresponds to the multiple sequence alignment comprising 1787 samples and the reference and only presenting the alignment informative sites (corresponding to the point 5 of the methodology)