Benchmarking pangenome graph mapping with strobemer-based seeding
Résumé
A pangenome represents the total genetic diversity of a species or a species complex. A recent model, the variation graph, aims to integrate full length genomes. Contrary to typical mapping approaches which rely on a single reference genome and are intrinsically biased, mapping query sequences to the full genetic variability of a species leads to more accurate mappings and improved genotype /phenotype analyses.
Mapping to a graph is similar to classic genome mapping and the first step involves a seed & extend approach, to target regions in the graph involving similar sequences before the steps of path
selection and alignment. When divergent query sequences are analyzed, this step can be problematic : not enough k-mer seeds are detected, and the following steps do not happen.
This work aims to test the potential of strobemers, an alternative to classic k-mer+minimizer seeding that allows more sequence divergence, in the context of mapping to pangenome graphs.
Domaines
Bio-informatique [q-bio.QM]Origine | Fichiers produits par l'(les) auteur(s) |
---|