Sample-specific haplotype-resolved protein isoform characterization via long-read RNA-seq-based proteogenomics

Area of Science:

• Proteogenomics
• Molecular Biology
• Bioinformatics

Background:

• Mass spectrometry (MS)-based proteogenomics relies on accurate protein databases.
• Existing methods often fail to capture the full genetic and transcriptomic diversity within a sample.
• Long-read RNA sequencing (lrRNA-seq) offers a way to identify complete transcript structures and genetic variants simultaneously.

Purpose of the Study:

• To develop an end-to-end workflow for constructing and searching haplotype-resolved, sample-specific proteomes.
• To leverage matched lrRNA-seq and MS data for improved proteogenomics.
• To enable the detection of allele-specific protein isoforms and linked variants.

Main Methods:

• Benchmarking phasing algorithms on PacBio lrRNA-seq data.
• Developing a Snakemake pipeline for variant calling, read-based phasing, transcript discovery, and proteome construction.
• Searching sample-specific, haplotype-resolved proteomes against MS data.

Main Results:

• Identified high-accuracy and complete phasing methods for lrRNA-seq.
• Demonstrated that sample-specific databases detect variant and splice peptides, allele-specific isoforms, and linked variants missed by reference proteomes.
• Successfully applied the workflow to cell lines and differentiation time courses.

Conclusions:

• lrRNA-seq-based phasing is a feasible and effective approach for proteogenomics.
• The developed workflow provides a practical framework for allele-resolved proteome characterization.
• This method enhances the characterization of proteomes in dynamic or disease-relevant biological settings.