Fig. 2
From: KAGE: fast alignment-free graph-based genotyping of SNPs and short indels
Graph with duplicate kmers. The SNP G/T has a kmer CTA (green) on the variant allele that also exists on the reference path (blue). If we observe each of the kmers CTA and CGA once in the read set, we might be fooled to believe that GT is the most likely genotype for this variant. However, when knowing that the kmer CTA is expected to occur at least once in the read data set, due to the duplication, we might conclude otherwise. This information can be used to adjust the probabilities used to compute the binomial probabilities of observing kmer counts given genotypes