Preprint Open Access

Comparison of Oxford Nanopore basecalling tools

Ryan Wick; Louise M Judd; Kathryn E Holt

This repository uses a bacterial genome to assess the read accuracy and consensus sequence accuracy for Oxford Nanopore Technologies (ONT) basecallers. Albacore v2.1.10, Guppy v0.5.1 and Scrappie raw v1.3.0 (all developed by ONT) were the best performers for read accuracy, and Chiron v0.3 produced the best assemblies. Consensus sequence accuracies reached approximately 99.75%, revealing that even the best basecallers still have systematic error. Nanopolish, used with its methylation-aware option, was able to raise consensus accuracy to about 99.9%. Most post-Nanopolish assemblies have similar accuracy, making basecaller choice relatively unimportant if Nanopolish is used.

Files (77.9 MB)
Name Size
rrwick/Basecalling-comparison-v5.1.zip
md5:c8d1b4647929749af905b7f57365df9a
77.9 MB Download
2,979
296
views
downloads
All versions This version
Views 2,9791,949
Downloads 296234
Data volume 22.3 GB18.2 GB
Unique views 2,6051,739
Unique downloads 252198

Share

Cite as