Fig. 3: Transmission inference across diverse taxonomic groups using TRACS.

a, A SARS-CoV-2 transmission network comprising samples that contain multiple distinct strains. Solid lines indicate transmission events observable using both TRACS and the typical consensus-based sequencing, while dashed lines represent additional links identified by TRACS alone. b, The inferred SNP distance between 21 pairs of 39 samples containing the same strain, where at least 1 sample includes multiple distinct strains. The consensus approach considers the dominant allele across the entire SARS-CoV-2 reference genome, whereas the ‘original’ approach excludes sites frequently filtered out to avoid hypervariable or error-prone regions. The TRACS algorithm correctly infers zero SNPs in all pairs. Box plots indicate the distribution of the inferred number of SNPS. The central line represents the median, the edges of the box indicate the IQR and the whiskers extend to 1.5 times the IQR. c, The expected number of intermediate hosts between each pair of 3,761 pneumococcal samples taken from different individuals in the Maela refugee camp versus the geographic distance between their homes within the camp with a LOESS smoothing. The shaded area indicates the corresponding confidence interval. Box plots show the median expected number of intermediate hosts. The edges of each box indicate the IQR, and the whiskers extend to 1.5 times the IQR. Sample pairs with a divergence time outside the establishment of the camp in 1984 were excluded. d, Similar to c, but only distances involving the three most common GPSCs found in 796 samples are shown. A strong geographic signal is absent for the non-multi-drug resistant lineage GPSC 1, which is known to have a longer carriage duration. e,f, The distribution of SNP distances inferred by TRACS for bulk versus single cell (e) and single cell versus single cell (f) samples. The vertical red line indicates the SNP threshold inferred using the TRACS mixture distribution approach. The high number of short SNP distances between single cells from the same sample indicates that TRACS can accurately distinguish closely related genomes within mixed infections of P. falciparum.