PING

PING infers nucleosome positions from MNase-Seq and MNase- or sonicated-ChIP-Seq data using an empirical Bayes mixture model to provide robust, scalable nucleosome-positioning analysis.

Key Features:

• Model-Based Methodology: Uses a probabilistic empirical Bayes mixture-model framework to analyze sequencing read densities and infer nucleosome positions, enabling robust inference at low read density.
• Scalability and Accuracy: Provides superior scalability compared to NPS (Nucleosome Positioning Software) and TemplateFilter while maintaining positional accuracy on large datasets.
• Robustness: Maintains high performance across varying sequencing data conditions.
• High-resolution Spatial Predictions: Produces high-resolution spatial nucleosome-position predictions suitable for biological inference of chromatin structure.

Scientific Applications:

• Detection of Nucleosome Positioning Changes: Detects alterations in nucleosome positioning around transcription factor binding sites, demonstrated using Illumina H3K4me1 ChIP-seq data under tamoxifen stimulation.
• Discrimination of Transcription Factor Binding Sites: Distinguishes functional from non-functional transcription factor binding sites more effectively than traditional enrichment profile analyses.
• Validation of Pioneer Transcription Factors: Supports validation of pioneer transcription factor associations, for example linking Foxa2 to the accessible major groove of nucleosomal DNA.

Methodology:

Analyzes MNase-Seq and MNase- or sonicated-ChIP-Seq read densities by fitting an empirical Bayes mixture model that compares observed read densities to probabilistic expectations to infer high-resolution nucleosome positions.