Skip to main content
Figure 2 | BMC Bioinformatics

Figure 2

From: HTJoinSolver: Human immunoglobulin VDJ partitioning using approximate dynamic programming constrained by conserved motifs

Figure 2

Matrix calculation of the alignment score for a sequence with a mutation or indel. Matrix calculation of the alignment score for a sequence with a mutation or indel. (a) Matrix for a single nucleotide mismatch. (b) Matrix with a two-base insertion (CG > CAAG). (c) Matrix with a two-base deletion (TC > −−). The dynamic programming matrix for the approximate backwards algorithm begins at the initial T of the VH-motif (last row and column, score = 0). The algorithm goes backwards along the diagonal until it hits a mismatch, in which the algorithm backs up a step and generates a submatrix (solid lines). The algorithm can choose to step up (deletion), step to the left (insertion), or continue diagonally (match/mismatch). For a deletion or insertion, the score initially decrease by 10, but subsequent indels have a score decrease of 4. Matches increase the score by 5, and mismatches decrease the score by 4. The maximum score in the first column or row (bold box) is selected (circled). The algorithm continues stepping backwards on the diagonal. Backtraces are shown as arrows, and label the alignment of the sequences.

Back to article page