LIQA

LIQA quantifies isoform-specific expression from long-read RNA sequencing (RNA-seq) data, accounting for high per-base error rates, chimeric reads, and alternative alignments.

Key Features:

• Expectation-Maximization Algorithm: Employs an Expectation-Maximization (EM) statistical method tailored for long-read data to estimate isoform abundances and support detection of differential alternative splicing (DAS).
• Read Quality and Length Integration: Integrates base-pair quality scores and isoform-specific read length information to assign weights to reads based on alignment confidence for improved quantification.
• Differential Alternative Splicing Detection: Identifies differential alternative splicing (DAS) events by leveraging isoform usage estimates derived from long-read data.
• Performance and Validation: Evaluated on simulated datasets and validated on Oxford Nanopore direct mRNA and cDNA sequencing datasets, paired short-read RNA-seq, and qPCR, showing improved characterization of low-coverage isoforms and DAS detection versus existing approaches.
• Application to PacBio Data: Applied to a PacBio dataset from esophageal squamous epithelial cells to identify DAS events not detected in corresponding short-read datasets.

Scientific Applications:

• Isoform Discovery: Accurately quantifies full-length transcripts to facilitate discovery of novel isoforms and their relative abundances.
• Alternative Splicing Analysis: Enables investigation of alternative splicing mechanisms and detection of condition-specific splicing changes via DAS analysis.
• Comparative Transcriptomics: Supports comparative studies of transcriptomic profiles across conditions to identify condition-specific isoform usage.

Methodology:

Processes long-read RNA-seq data in BAM format with a reference isoform annotation file, applies an EM algorithm, and integrates base-pair quality scores and isoform-specific read length information to weight reads by alignment confidence.