nanopolish

Nanopolish is an open-source software package for the analysis of nanopore sequencing data produced by Oxford Nanopore Technologies. It focuses on analyzing the raw electrical signal captured as reads pass through nanopores, stored in fast5 files, in conjunction with basecalled reads and sequence alignments. The program is used to improve consensus accuracy, call variants, and detect DNA methylation and other base modifications from nanopore data.

Key capabilities include polishing draft assemblies by reinterpreting signal-level data to correct errors, calling single-nucleotide polymorphisms and small insertions or deletions from aligned reads, and identifying modified bases by comparing observed signal patterns to unmodified and modified models. The methylation module is designed to detect 5-methylcytosine in CpG contexts, among others.

Typical workflow involves basecalling with a separate basecaller, aligning reads to a reference or draft assembly, indexing the fast5 files, and running nanopolish with commands such as eventalign, variants, and call-methylation. The eventalign step maps raw signal events to the reference, providing data for downstream variant or methylation analysis. The variants module produces VCF-like calls, while the methylation module outputs per-site methylation probabilities.

Nanopolish is distributed as open-source software and is widely used in genomics research. It runs on common Unix-like systems and integrates with other standard bioinformatics tools. The project is hosted on a public repository and maintained by contributions from the research community.