FracFixR

FracFixR reconstructs original fraction proportions in RNA sequencing data by modeling the compositional relationship between whole and fractionated RNA samples, correcting for bias introduced by library preparation and sequencing depth.

Key Features:

• Global Fraction Weight Estimation: Uses non-negative linear regression on carefully selected transcripts to estimate global fraction weights.
• Transcript Frequency Correction: Corrects individual transcript frequencies based on estimated fraction weights.
• Lost Fraction Quantification: Quantifies unrecoverable material that is lost during fractionation.
• Differential Proportion Testing: Includes methods for differential proportion testing between conditions using binomial GLM, logit, or beta-binomial models.

Scientific Applications:

• RNA Localization Studies: Enables accurate interpretation of fractionated RNA-seq data used to study RNA localization and subcellular compartmentalization.
• Translation Analysis: Supports analysis of polysome profiling data to detect differentially translated transcripts between conditions such as cancer subtypes.
• Compositional Bias Correction: Corrects for systematic bias that compromises comparisons of fraction-specific RNA profiles when biological changes shift RNA distribution across fractions.

Methodology:

FracFixR models the compositional relationship between whole and fractionated RNA samples using non-negative linear regression on carefully selected transcripts to estimate global fraction weights, correct individual transcript frequencies, and quantify unrecoverable material. Differential proportion testing between conditions is performed using binomial GLM, logit, or beta-binomial models.