By using data comparisons of known proteins or protein subunits stored in data bases, the structure and function of related proteins can be predicted. Until now researchers have used such tools as the standard software PSI-BLAST. The new software HHblits (Homology detection by iterative HMM-HMM comparison) analyses the sequence of the molecular building blocks that make up proteins, the amino acid sequence, more than 2,500 times faster than the program PSI-BLAST.

The development of HHblits is considerably superior to all data base sequence comparisons using the standard program PSI-BLAST. To achieve this, the bioinformaticians transform the sequences to be analysed and the comparison sequences in the data bases into so-called Hidden Markov models (HMM). HMMs are statistical models of amino acid sequences which also take the probability of mutations which can be read out of the sequence alignments into account, leading to a more sensitive and more precise search.

The bioinformatics scientists intend to expand their method by taking into account structural information of the proteins. The researchers published their method in the journal Nature Methods (2011, online publication).

More information:

Sources:
http://www.biotechnologie.de/BIO/Navigation/DE/root,did=147480.html
Abstract Nature Methods: http://www.nature.com/nmeth/journal/vaop/ncurrent/full/nmeth.1818.html
LMU: http://www.genzentrum.lmu.de/ueber-das-genzentrum
HHblits: http://toolkit.genzentrum.lmu.de/hhblits