Analysing long-read CRISPR experiments with CRISPRLungo.

Long-read amplicon sequencing analysis combining unique molecular identifier-based error correction with statistical filtering to tell true editing events (indels, inversions) from background noise, including at the SBDS locus despite the homologous SBDSP1 pseudogene.

CRISPRLungo is a computational pipeline built for long-read amplicon sequencing of gene-edited samples. It combines unique molecular identifier-based error correction with statistical filtering to separate genuine editing events from background noise, enabling robust detection of both small indels and structural variants. Systematic benchmarking on simulated datasets shows it outperforms existing approaches in accuracy and read recovery; it supports Oxford Nanopore and PacBio and uncovered previously undetected structural edits such as inversions in published CRISPR datasets. Applied to edited primary cells from a patient carrying compound heterozygous SBDS mutations, it quantified allele-specific editing outcomes despite contaminating reads from the homologous SBDSP1 pseudogene. The pipeline is freely available on GitHub together with a fully client-side web application that requires no installation.

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

The contribution is methodological: the evidence rests on simulations and a single patient sample, with no orthogonal confirmation of the detected events, so the accuracy gains remain to be tested on annotated real-world data. For clinical genetics the tool sits upstream — quality control of editing experiments — and does not change how patient variants are interpreted. Its explicit handling of the SBDSP1 pseudogene is, however, a practical gain at any locus with high sequence homology.

Analysis by Dr Thibaut Benquey