Figure 6. PE/PPE genes vary considerably within host while putative antigens remain conserved.

(A) Mutational spectrum of in-host SNPs. (B) In-host SNP counts vs. time between isolate collection (195/200 patients with dates shown, *W [Walker et al., 2013] isolates only had year of collection). (C) Boxplots of nucleotide diversity by gene within each of five non-redundant categories (see text; $n=$ number of genes). (D) Average nucleotide diversity across genes by category. Nucleotide diversity in epitope and non-epitope region (Materials and methods) of each gene in the Antigen (E, F) and PE/PPE (G, H) gene categories. (l, J) PE/PPE genes separated into three non-redundant categories: PE, PE-PGRS, and PPE. (J) The average nucleotide diversity by category. (I) Box plot of nucleotide diversity by gene.

Figure 6—figure supplement 1. Basic characteristics of epitopes used in analysis.

We downloaded a set of 2031 epitope peptide sequences from IEDB (Vita et al., 2015) and used BLASTP to map these peptide sequences to H37Rv imposing an e-value cut-off of 0.01 (Materials and methods). (A) The distribution of e-values and (B) distribution of peptide lengths for the retained epitope mappings.

Figure 6—figure supplement 2. Most T cell epitopes remain conserved in-host during active TB disease.

No SNPs were detected in-host for a vast majority of CD4⁺ and CD8⁺ T cell epitopes, however, 1 SNP was detected in a small number $\left(n=5\right)$ of overlapping epitopes in PPE18. A list of these epitopes is given in Supplementary file 12.