HHsenser: exhaustive transitive profile search using HMM-HMM comparison

Area of Science:

• Bioinformatics
• Computational Biology
• Protein Sequence Analysis

Background:

• Identifying homologous proteins is crucial for understanding protein function and evolution.
• Existing methods may struggle with sensitivity or produce false positives, especially for divergent sequences.
• Accurate multiple sequence alignments are essential for downstream applications like structure and function prediction.

Purpose of the Study:

• To introduce HHsenser, a new computational tool for sensitive and accurate detection of protein homologs.
• To provide a method for exploring protein superfamilies starting from a single sequence or alignment.
• To improve the accuracy of homology-based predictions, particularly for challenging sequences.

Main Methods:

• Utilizes exhaustive intermediate profile searches combined with pairwise comparison of hidden Markov models (HMMs).
• Employs an iterative approach to explore entire protein superfamilies.
• Generates multiple sequence alignments of detected homologs as output.

Main Results:

• HHsenser achieves high sensitivity with minimal false positives in identifying homologous sequences.
• The tool can effectively explore protein superfamilies from limited input data.
• The generated multiple alignments are suitable for various downstream analyses.

Conclusions:

• HHsenser is a powerful and sensitive tool for protein homology detection and evolutionary studies.
• Its ability to produce accurate alignments enhances applications in structure and function prediction.
• Integration into HHpred improves predictions for difficult-to-analyze sequences.