From: Koonin, Eugene (NIH/NLM/NCBI) [E]
Sent: Tuesday, March 03, 2009 1:22 PM
To: kalinina@embl.de; NLM/NCBI List ncbi-seminar
Subject: special seminar March 23

Monday, March 23, 11 AM

Bldg. 38A, B2 Conference Room

 

Olga Kalinina

 

European Molecular Biology Laboratory (EMBL), Heidelberg, Germany

Studying protein sequence alignments: specificity determinants, functionally important sites and correlated mutations

The rapid expansion of sequence databases allows for new methodologies in the sequence analysis field. We use statistical
 properties of protein multiple sequence alignments to predict the role that certain residues may play in the function of 
the protein. Particularly, comparing of groups of proteins with similar functional specificity within same protein family 
one can identify positions that correlate with these groups, and thus probably account for the differences in specificity.
 We demonstrate this approach to be efficient in identification of probable specificity determinants. It can be extended 
to contribute to several long-standing bioinformatics problems, such as annotation of protein specificity, prediction of 
functional site position, enzyme activity and protein-protein interactions.