USeq

USeq provides analysis of next-generation, ultra-high-throughput signature sequencing data to detect genomic features and estimate confidence in ChIP-seq and RNA-Seq experiments generated on Solexa, SOLiD, and 454 platforms.

Key Features:

• False Discovery Rate Estimation: Incorporates algorithms to estimate the False Discovery Rate (FDR) for peak detection in sequencing data.
• Control Input Data Utilization: Leverages control input data to reduce false positives and improve ChIP-Seq peak recovery, reported to more than double peak recovery at 5% FDR.
• Confidence Estimation Methods: Implements binomial p-value/q-value and empirical FDR approaches that predict true FDR within 2-3 fold and are more reliable than global Poisson p-values.
• Independence from Prior Knowledge: Estimates peaks and confidence without requiring prior knowledge of ChIP targets or validated binding sites.

Scientific Applications:

• Transcription Factor Binding Site Identification: Identifies transcription factor binding sites from chromatin immunoprecipitation (ChIP) data using signature sequencing mapping.
• Histone Modification and Chromatin Structure Characterization: Maps histone modifications and chromatin structure changes genome-wide.
• DNA Methylation Pattern Profiling: Profiles DNA methylation patterns across genomes.
• Gene Regulation and Chromatin Remodeling Analysis: Supports reconstruction of gene regulation networks and analysis of chromatin remodeling from enriched sequence mappings.
• Biological Context Studies: Enables studies of developmental biology, cellular responses to perturbations, and neoplastic transformations.

Methodology:

Compares methods using simulated spike-in datasets, uses control input data and normalized difference scores to enhance peak recovery, and reanalyzes existing ChIP-Seq datasets such as Johnson et al.'s neuron-restrictive silencer factor (NRSF) data.