The Single Read SNV (SRSNV) pipeline is a read-centric de-noising framework, developed to overcome the limitations of traditional locus-centric variant calling, particularly in scenarios where rare mutations may be supported by only a single read. These rare mutations need to be distinguished from artefactual SNVs, which can derive from sequencing, library or alignment errors. To achieve this, we employed a supervised machine learning model trained to classify actual SNVs (labelled True or TP) f
This is a companion discussion topic for the original entry at github.com/bicklab/healthomics-workflows-single-read-snv/SingleReadSNV