motifRG

motifRG applies regression-based discriminative motif discovery to identify transcription factor (TF) binding motifs that differ between two sequence datasets from high-throughput assays such as ChIP-seq and DNase accessibility, improving biological relevance.

Key Features:

• Differential motif discovery: Discovers motifs that differentiate between two sets of sequences to resolve TF-specific binding differences.
• Regression-based identification: Employs regression methods to detect differential motifs between positive and background sequence datasets.
• Bias elimination and scalability: Incorporates an algorithm to eliminate systematic biases and scales to large high-throughput datasets such as ENCODE ChIP-seq.
• Performance on ENCODE datasets: Evaluated on 207 ENCODE ChIP-seq datasets and correctly identified known motifs in 78% of cases, outperforming DREME in benchmarks.
• Handling complex datasets: Detects underlying technical or biological factors that can compromise motif searches and enables control for these confounders.
• Single base-pair resolution: Detects single base pair differences in DNA specificity between similar transcription factors.
• Application to tissue specification: Applied to high-throughput DNase accessibility data to discover key TF motifs involved in tissue specification.

Scientific Applications:

• Transcription factor binding analysis: Identification and comparison of TF binding motifs from ChIP-seq and DNase accessibility datasets.
• Gene regulation studies: Characterization of motif contributions to regulatory mechanisms and TF specificity.
• Tissue-specific regulatory motif discovery: Discovery of TF motifs underlying tissue specification using DNase accessibility data.
• Comparative motif specificity: Resolving single base pair differences in DNA recognition between closely related TFs.
• Large-scale dataset analysis: Analysis and bias-controlled motif discovery in large consortia datasets such as ENCODE.

Methodology:

Uses regression methods for discriminative motif discovery and an algorithm to eliminate systematic biases when identifying differential motifs between two sequence datasets.