PGpath: recommending population-adapted linear reference genomes from pangenome subpaths.

PGpath generates population-adapted linear reference genomes from pangenome subpaths, improving read mapping and variant detection accuracy in large-scale genomic analysis.

• Why it matters: Accurate representation of genetic diversity across populations is essential for reliable genomic analysis, but current references are insufficient and pangenome graphs are complex and incompatible with standard workflows.
• What they did: The study developed PGpath, a method that uses population-level k-mer features and a topology-constrained neural network to recommend linear references from pangenome subpaths, tested on HPRC datasets and HG002 benchmarks.
• The result: PGpath-derived references enhanced read mapping performance and decreased variant detection errors, offering a practical approach to incorporate pangenome diversity into population-specific linear references.