Estimating protein isoform abundances with PAQu.

Area of Science:

• Proteomics
• Genomics
• Bioinformatics

Background:

• Genes encode multiple protein isoforms with distinct functions, crucial for biological processes.
• Current methods struggle to quantify protein isoform abundance accurately, especially with ambiguous peptide mapping.
• Transcript levels alone do not fully explain cellular control of isoform abundances.

Purpose of the Study:

• To develop a novel computational method for accurate, large-scale quantification of protein isoform abundance.
• To integrate multiomic data (peptidome and transcriptome) for improved isoform quantification.
• To enable robust hypothesis testing for differential isoform expression.

Main Methods:

• Introduction of PAQu, a Bayesian statistical framework.
• Leveraging multiomic information from peptidome and transcriptome data.
• Utilizing uncertainty quantification and rigorous hypothesis testing.

Main Results:

• PAQu accurately estimates isoform abundance, even with ambiguous peptide-to-isoform mapping.
• Simulations demonstrate PAQu's superior performance over existing methods in detecting differential isoform expression.
• Application to schizophrenia data confirmed increased C4A and unchanged C4B isoform levels, validating a long-standing hypothesis.

Conclusions:

• PAQu provides a powerful, unified framework for accurate protein isoform quantification.
• The method enables the identification of significant isoform abundance variations previously undetectable.
• PAQu has implications for understanding molecular mechanisms in diseases like schizophrenia.