The task

Goal

Produce a program in Python programming language, which given read alignments (in BAM format with haplotype tags) and a set of variants (in phased VCF format) computes support for ALT and REF alleles across reads assigned to individual haplotypes.

Note that usage of GenAI coding assistants (GitHub Copilot, Claude Code, etc) is both allowed and encouraged.

Inputs & assumptions

• alignments.bam -- BAM file with read alignments. Coordinate-sorted, indexed (matching .bai index is also provided) alignments BAM, with primary alignments optionally haplotagged withHPtag (1 or 2).
• variants.vcf(.gz) -- variants in phased VCF format (optionally gzipped). To keep things relatively simple, we only ask you to consider bi-allelic mismatch variants (SNVs), ignoring indels, MNPs, etc.

Output

Produce a TSV file with rows corresponding to SNVs and the following 6 required columns (feel free to include extra columns if you think they might be useful):

• chrom – chromosome name
• pos– position on chromosome
• h<H>_<A> with H in {1,2} and A in {'ALT, 'REF'} (4 columns total) — number of primary alignments with HP==H supporting allele A

Test data

Test data archive contains read alignments (30x coverage) and variants for chr16:28000000-28500000 region of HG002 genome.