HH-MOTiF

HH-MOTiF identifies de novo short linear motifs (SLiMs) in protein sequences by constructing and comparing Hidden Markov Models (HMMs) from input proteins and their closely related orthologs to detect short stretches of homology.

Key Features:

• Evolutionary information: Incorporates evolutionary information by building HMMs from each input protein and its closely related orthologs.
• HMM construction: Constructs Hidden Markov Models (HMMs) for each input sequence and its orthologs.
• HMM comparison: Compares HMMs against one another to identify short stretches of homology indicative of SLiMs.
• Hierarchical motif representation: Represents identified motifs as hierarchical motif trees to capture relationships among motifs.
• Degenerate motif detection: Enables identification of degenerate motifs while maintaining high precision.
• Performance balance: Balances recall and precision and reports improved performance versus other methods on test data.

Scientific Applications:

• De novo SLiM discovery: Prediction of short linear motifs (SLiMs) across sets of primarily unrelated proteins.
• Protein interaction studies: Identification of motif-mediated interaction sites within proteins.
• Computational and experimental research: Support for computational analyses and experimental validation of motif functions.

Methodology:

Builds HMMs from each input protein and its closely related orthologs, compares these HMMs to detect short stretches of homology as candidate SLiMs, and organizes identified motifs into hierarchical motif trees.