Researchers have introduced longcallR, a new computational tool designed to analyze long-read RNA sequencing data by integrating SNP calling, haplotype phasing, and allele-specific analysis. This pipeline addresses the unique difficulties of transcriptomic data, such as uneven coverage and alignment errors, by using a deep convolutional neural network to predict genetic variants. Beyond identifying mutations, the software group reads into specific parental haplotypes to detect allele-specific expression and splicing events with high precision. When tested on over 200 human samples, the tool revealed numerous allele-specific junctions, many of which involved previously undocumented genetic structures. By linking genetic variation directly to transcript diversity, this method provides a more comprehensive view of how individual genomes influence gene regulation and potential disease mechanisms.

References:

• Huang N, Li H. SNP calling, haplotype phasing and allele-specific analysis with long RNA-seq reads[J]. Nature Methods, 2026: 1-6.