Fig. 1

Variant-aware transcript detection by FLAIR2 identifies haplotype-specific transcript isoform bias. a Full FLAIR2 computational workflow for identifying haplotype-specific transcripts in long reads. For annotated transcript discovery, long reads are aligned to annotated transcript sequences and inspected for their overall match and read support at annotated splice junctions and transcript ends. The genomic alignments for reads that are not assigned to an annotated transcript are corrected and collapsed for unannotated isoform discovery. User-provided unphased/phased RNA variant calls can be associated with reads using FLAIR2; last, FLAIR2 counts the number of variant sets comprised by the reads assigned to each transcript model to determine variant-aware transcripts. Red ticks indicate mismatches; purple stars indicate RNA variants. b FLAIR transcript models for Mcm5 with the highest expression are plotted using different colors for each transcript’s exons. The highlighted portion shows alternative splicing and the smaller blocks within exons indicate variants. c Stacked bar chart showing the proportion of transcript expression of transcripts from b as matched by color for each of the replicates sequenced