BCIG TUTORIAL: "Semi-Global and Local Alignment of Protein Domains"

Clinical Center (Building 10) Old Medical Board Room (Room 2C116)

(view presentation)

DESCRIPTION: With complete genome sequencing now routine, biology faces the fundamental problem of large-scale automatic annotation of gene function. The most powerful approach for inferring function of new protein sequences is the transfer of annotation from similar sequences using sequence comparison methods. Because proteins are composed of basic units called domains, a gene can be annotated using a domain database by aligning domains to the gene's protein sequence. In this talk, I will briefly introduce the different methodologies to search for domains within proteins. These search alignment tools can be classified, depending on the extent of the sequences aligned, as local, semi-global, and global. Given the modular, multi-domain nature of proteins, local alignments are indispensable tools since they detect similarities involving only one or a subset of domains. However, since intact protein domains are the basic units of protein folding, evolution, and function, a semi-global alignment tool is useful when comparing protein queries to a database of protein domain sequences or search models, and it may be required, if one was to certify the presence of complete domains. In the past, the lack of accurate statistics for expected scores generated by the semi-global alignment tools has hampered studies in this direction. I will introduce a new semi-global alignment tool that we have developed at NCBI. Contrary to other semi-global tools, this new method, GLOBAL, provides extremely accurate score statistics. When local and semi-global methods were benchmarked using a structure-based standard of truth, we found that when searching in a protein domain database, semi-global alignment methods retrieve domains significantly better than the local alignment tools. The heuristic acceleration used in other local alignment tools can be implemented into GLOBAL to further increase the search speed, making it an ideal tool for high-throughput protein domain searches with accurate p-values.

REGISTRATION: As with all BCIG events, registration is not required. Just show up happy.

WEBCASTING: This event will be web cast live and be made available for post program viewing on the BCIG web site (www.nih-bcig.org). To get more information about our webcasting service, please contact Meeting Master Carl Leonard by e-mail: cleonard@lired.com or by calling him on 301-496-0191. NIH CONTACT: Jim DeLeo, 301-496-3848, jdeleo@nih.gov

REFRESHMENTS: Bring refreshments if you would like. There is an open cafeteria near the meeting room.

BCIG WEB SITE: www.nih-bcig.org

NIH VISITOR INFORMATION: http://www.nih.gov/about/visitor/ 

3:00 - 4:30 pm November 9, 2006

Maricel G. Kann, Ph.D., (Postdoctoral Fellow) National Center for Biotechnology Information (NCBI), NIH

Related Links