DRIMust

DRIMust identifies motifs in ranked sequence lists using suffix trees and a minimum hypergeometric statistical framework to detect over-represented variable gapped, long, and large-alphabet motifs and to report P-values and Position Specific Scoring Matrices (PSSMs).

Key Features:

• Variable Gapped Motifs: Identifies motifs comprising two half-sites separated by a flexible-length gap.
• Long Motifs and Large Alphabets: Searches for long motifs across large alphabets to accommodate complex biological sequences.
• Efficient Enumeration with Suffix Trees: Utilizes suffix trees to efficiently enumerate motif candidates.
• Minimum Hypergeometric Statistical Framework: Applies the minimum hypergeometric statistic to assess motif enrichment rigorously.
• Data-Driven Thresholds: Dynamically determines thresholds for identifying over-represented motifs at the top of ranked lists.
• Accurate P-value Assessment: Computes precise P-values for detected motifs.
• Position Specific Scoring Matrix (PSSM): Outputs discovered motifs as individual motifs with associated P-values and as PSSMs.

Scientific Applications:

• Sequence-based recognition (DNA, RNA, proteins): Supports study of sequence-based recognition processes involved in molecular regulation and signaling across DNA, RNA, and proteins.
• High-throughput dataset analysis: Applied to analyze high-throughput measurement datasets for enriched sequence motifs.
• Motif refinement and novel motif discovery: Used to refine known motifs such as the human estrogen receptor 1 motif and to identify variable-length motifs potentially associated with tyrosine phosphorylation.

Methodology:

Processes ranked lists of sequences in FASTA format; enumerates motif candidates using suffix trees; assesses motif significance with the minimum hypergeometric statistic; determines data-driven thresholds for enrichment; computes motif P-values and outputs motifs with Position Specific Scoring Matrices (PSSMs).