MACPET

MACPET analyzes paired-end ChIA-PET sequencing reads to identify protein-DNA binding sites and associated three-dimensional genomic interactions.

Key Features:

• Input formats: Processes ChIA-PET data in BAM and SAM formats.
• PET classification: Categorizes paired-end tags (PETs) into Self-ligated, Intra-chromosomal, and Inter-chromosomal classes.
• Genome segmentation and modeling: Divides the genome into regions and applies two-dimensional mixture models to identify candidate peaks or binding sites.
• Distributional modeling: Models signal and noise using skewed generalized students-t distributions (SGT) within the mixture models.
• Statistical significance: Employs a local Poisson model to determine statistically significant binding sites.
• Comparative performance: Shows improved motif occurrence, spatial resolution, and false discovery rate relative to MACS.
• 3D interaction linking: Links discovered binding sites with a higher number of significant 3D genomic interactions.

Scientific Applications:

• Protein binding site identification: Detecting protein-DNA binding sites from ChIA-PET data.
• Chromatin architecture: Elucidating three-dimensional genomic interactions and chromatin organization.
• Gene regulation studies: Investigating relationships between binding sites and regulatory activity relevant to gene regulation and chromatin dynamics.

Methodology:

Processes paired-end reads in BAM/SAM, classifies PETs into Self-ligated, Intra-chromosomal, and Inter-chromosomal categories, divides the genome into regions, applies two-dimensional mixture models using skewed generalized students-t distributions (SGT), and uses a local Poisson model to assess binding site significance.