Fig. 2From: A sensitive short read homology search tool for paired-end read sequencing dataHMM alignments of a read pair. Paired-end reads r 1 and r 2 represented by two greyscale lines are aligned against models M 1, M 2, and M 3 with different scores of alignments. The darker lines represent bigger scores. The fragment size distribution is provided above each model. The distance between the two alignments is computed and is used to compute the likelihood of the corresponding fragment size. In this example, M 1 is most likely to be the native familyBack to article page