BBUM

BBUM models differential expression from small-RNA sequencing to distinguish primary microRNA substrates of ZSWIM8-mediated target-directed microRNA degradation (TDMD) from secondary effects and experimental noise.

Key Features:

• Bi-Beta-Uniform Mixture Model: Implements a bi-beta-uniform mixture (BBUM) statistical model to separate primary signal components from background noise in differential-expression data.
• Directional focus on upregulation: Leverages the expectation that primary substrates are upregulated upon loss of ZSWIM8 to restrict detection to the relevant direction of change.
• Robustness and consistency: Provides robustness to outliers and maintains consistent stringency across analyses using a single false discovery rate (FDR) cutoff.
• Statistical rigor: Incorporates FDR-correction and significance-testing methods to control false positives when identifying primary substrates.

Scientific Applications:

• Identification of ZSWIM8 primary substrates: Detects microRNAs directly targeted by ZSWIM8-mediated TDMD in small-RNA-seq differential-expression experiments.
• Discrimination of direct versus secondary effects: Separates primary regulatory targets from secondary expression changes in studies of regulatory factors.
• Adaptation to directional regulatory analyses: Applicable to scenarios where primary targets exhibit a specific directional change (upregulation or downregulation).

Methodology:

Fits a bi-beta-uniform mixture model to differential-expression data from small-RNA sequencing, models the distribution of signal components to isolate primary signals, and applies FDR correction and significance testing.