Perplexity as a Metric for Isoform Diversity in the Human Transcriptome.

Area of Science:

• Genomics
• Molecular Biology
• Bioinformatics

Background:

• Long-read sequencing (LRS) reveals extensive RNA isoform diversity.
• Existing analysis methods use arbitrary expression cutoffs, limiting biological interpretation.
• A unified metric is needed to assess isoform complexity across diverse biological contexts.

Purpose of the Study:

• Introduce perplexity as a quantitative measure for RNA isoform diversity.
• Evaluate perplexity's utility in analyzing LRS data from human cell types.
• Establish perplexity as a standard for interpreting isoform expression.

Main Methods:

• Calculated perplexity from isoform ratio distributions in 124 ENCODE4 PacBio LRS datasets.
• Analyzed isoform diversity at transcript and protein (ORF) levels.
• Assessed tissue-specific expression patterns of ORFs.

Main Results:

• Perplexity effectively quantifies isoform diversity and uncertainty across genes.
• Average ORF-level perplexity is 2.1, indicating two distinct protein isoforms per gene.
• Identified 4,593 tissue-specific ORFs across 3,102 genes.

Conclusions:

• Perplexity provides an interpretable and consistent metric for RNA isoform diversity.
• This metric aids in understanding gene expression regulation across tissues and cell types.
• A community resource of perplexity analyses facilitates cross-study comparisons of novel isoforms.