ChIPanalyser

ChIPanalyser predicts and models transcription factor (TF) binding to DNA using a statistical thermodynamic framework to quantify effects of DNA sequence and accessibility.

Key Features:

• Explainability: Provides explainable models of TF–DNA interactions within a statistical thermodynamic framework.
• Versatility: Predicts TF binding across varied genomic contexts and chromatin states.
• Biological insights: Dissects how chromatin accessibility and sequence specificity shape TF binding profiles.
• Implementation: Implemented as an R/Bioconductor package.

Scientific Applications:

• CTCF: Prefers high-affinity sites primarily located within open chromatin in Drosophila cell lines.
• BEAF-32: Binds most of its high-affinity sites in open chromatin in Drosophila cell lines.
• su(Hw): Binds both open and partially closed chromatin in Drosophila cell lines.
• Ubx (Hox): Binds exclusively within open chromatin in Drosophila.
• Abd-B and Dfd (Hox): Capable of binding both open and partially closed chromatin in Drosophila.
• Chromatin accessibility versus TF concentration: Differences in TF binding profiles across cell lines are driven predominantly by variations in DNA accessibility rather than TF concentration.

Methodology:

Uses a statistical thermodynamic framework; scores potential binding sites with a Position Weight Matrix (PWM); incorporates DNA accessibility; accounts for the number of TFs bound and applies a binding specificity modulator to adjust binding energy or specificity.