Related Experiment Videos
The HHpred interactive server for protein homology detection and structure prediction
Johannes Söding1, Andreas Biegert, Andrei N Lupas
1Department of Protein Evolution, Max-Planck-Institute for Developmental Biology Spemannstrasse 35, 72076 Tübingen, Germany. johannes.soeding@tuebingen.mpg.de
Nucleic Acids Research
|June 28, 2005
Summary
HHpred is a fast server for remote protein homology detection and structure prediction. It uses profile hidden Markov models (HMMs) for efficient database searching and generates alignments and 3D models.
Area of Science:
- Bioinformatics
- Computational Biology
- Structural Biology
Background:
- Remote protein homology detection and structure prediction are crucial for understanding protein function.
- Existing methods often require significant computational resources and time.
Purpose of the Study:
- To introduce HHpred, a novel server for rapid remote protein homology detection and structure prediction.
- To implement pairwise comparison of profile hidden Markov models (HMMs) for enhanced accuracy.
Main Methods:
- Utilizes profile hidden Markov models (HMMs) for pairwise comparisons.
- Searches diverse biological databases including PDB, SCOP, Pfam, SMART, COGs, and CDD.
- Accepts single sequences or multiple alignments as input, providing results in a user-friendly format.
Main Results:
- HHpred achieves fast search speeds, delivering results comparable to PSI-BLAST within minutes.
- Enables local or global alignment, with options for scoring secondary structure similarity.
- Generates pairwise and multiple alignments, and constructs 3D structural models using MODELLER.
Conclusions:
- HHpred offers an efficient and user-friendly solution for protein homology detection and structure prediction.
- The server's ability to generate alignments and 3D models aids in functional and structural characterization of proteins.
- HHpred is a valuable tool for researchers in bioinformatics and structural biology.