NCM 2021: Visualisation and analysis of nanopore RNA and DNA signal alignments with UNCALLED4
- Home
- NCM 2021: Visualisation and analysis of nanopore RNA and DNA signal alignments with UNCALLED4
Several tools use nanopore signal data for detecting modifications, polishing references, calling variants, and real-time targeted sequencing. Many of these (nanopolish, Tombo, NanoSatellite, etc.) function by aligning the raw signal directly to a reference genome using a form of dynamic time warping (DTW), a dynamic programming alignment algorithm for time series data. There are many variations of DTW which affect accuracy and speed, and different tools use different implementations. Yet despite their widespread use, there is no such tool to assess the accuracy of nanopore signal alignments. In this breakout, Sam presented UNCALLED4: a toolkit for nanopore signal alignment and subsequent analysis. UNCALLED4 features interactive visualisations and command line tools for signal-to-reference alignments, comparisons between alignment methods, and a genome browser for signal alignment tracks. Comprehensive signal analyses can be performed and visualised, including modification detection, alignment sorting and clustering, creation of pore k-mer models, and more. Sam also introduced a novel DTW method that is guided by Guppy basecaller metadata, and enables the conversion of alignments from nanopolish and Tombo to their database format. Insights from these analyses have improved the rapid mapping mode for targeted sequencing introduced in UNCALLED version 1 and hold great potential for implementation to Oxford Nanopore’s adaptive sampling method.