GaussF: An Unbiased Gaussian Framework for High-Precision Transcript Quantification with a Programmable Self-Benchmark Model

GaussF enables unbiased, high-precision isoform quantification from short-read RNA-seq by using a transcript-specific theoretical Gaussian CDF benchmark with parameters independent of empirical sequencing data.

• Why it matters: Isoform-level quantification is hindered by overlapping isoform sequences and sequencing biases that current methods often correct using empirical, data-derived factors.
• What they did: GaussF models k-mer counts by GC content as an intrinsic Gaussian distribution, derives CDF parameters without empirical sequencing data, and improves resolution by using minimal-shared regions to restructure the k-mer reference.
• The result: Validated on artificial and real RNA samples, GaussF shows superior accurate isoform detection and quantification versus Kallisto, Salmon, Cufflinks, StringTie, and RSEM, and is released on GitHub and PyPI.

AI summary of the abstract · may contain errors · not medical advice · read the paper.