A De Novo Algorithm for Allele Reconstruction from Oxford Nanopore Amplicon Reads, with Application to CYP2D6.

A “sequence-first” algorithm that reconstructs allele sequences de novo from long amplicon reads (Oxford Nanopore), then compares them to the reference to infer the diplotype; target-gene agnostic, it detects novel variants and allele duplications.

The authors describe a generic algorithm that reconstructs allele sequences from long amplicon reads (Oxford Nanopore) using a “sequence-first” approach: rather than calling variants directly, it reconstructs the underlying sequences de novo, then compares them to the reference to infer the diplotype. Target-gene agnostic, it detects novel variants and allele duplications. Demonstrated on CYP2D6 (over 175 known alleles), it recovers validated diplotypes from 20 Coriell samples covering 14 distinct alleles, across amplicons, flow-cell versions and depths, and also extends to HLA. The code is open source (GitHub, Zenodo).

Synthesis written by Geno'X. For the full original abstract, please refer to the source publication.

A tool that tackles a real long-read bottleneck: turning raw reads into an actionable genotype without a gene-specific custom script. The gene-agnostic design and open-source code make it a reusable building block, notably for complex pharmacogenes such as CYP2D6. The demonstration remains proof-of-concept scale (20 Coriell samples): validation on larger clinical cohorts is needed before diagnostic deployment.

Analysis by Dr Thibaut Benquey