Overview of VILOCA. (A) VILOCA takes the aligned sequencing reads as input and proceeds through three main steps. (1) It defines the genome tiling and extracts the corresponding local alignments from the input alignment file. (2) For each local region, haplotypes are reconstructed using a clustering approach which is based on a Dirichlet process mixture model. In the graphical representation, each node represents a variable of the model; the shaded notes represent observed variables. (3) Finally, mutation calls across all regions are consolidated and subjected to a quality assessment. This assessment includes applying a strand bias filter for paired-end reads and evaluation of the posterior score. Subsequently, the tool reports local haplotypes and mutation calls that successfully pass the quality assessment. (B) Schematic illustration of two genome tiling approaches. In the case of uniform genome tiling, a sliding window method is utilized to distribute local regions of uniform length across the genome. Each position is covered by a number of user-defined local regions. Alternatively, when the primer strategy is known or can be specified, a bed file outlining the amplicons by start and end positions is used to define the regions.