BinQuasi

BinQuasi implements joint modeling of replicated ChIP-seq datasets to detect peaks representing DNA–protein interactions while controlling the false discovery rate.

Key Features:

• Joint Modeling Framework: Employs a generalized linear model framework to jointly analyze replicated ChIP-seq datasets.
• One-Sided Quasi-Likelihood Ratio Test: Uses a one-sided quasi-likelihood ratio test for peak detection and FDR control.
• Improved Performance and Reliability: Demonstrated superior peak classification accuracy and better FDR management on simulated and real-world ChIP-seq datasets.
• Flexibility in Analysis: Applies a flexible joint-modeling approach that accommodates the characteristics of biological replicates instead of combining replicate results post hoc.

Scientific Applications:

• Epigenetic Studies: Detects changes in chromatin-associated signals relevant to histone modifications and other epigenetic marks.
• Transcription Factor Binding Analysis: Identifies transcription factor binding sites across biological conditions or treatments using replicated ChIP-seq data.
• Comparative Genomics: Enables accurate comparative analyses of DNA–protein interaction profiles across species or cell types.

Methodology:

Jointly models biological replicates using a generalized linear model framework and applies a one-sided quasi-likelihood ratio test for peak detection; performance and FDR control were evaluated on simulated and real-world ChIP-seq datasets.