Seminar notice: Bill Pearson, Univ. of VA School of Medicine Friday, Nov. 14, 1:00 BH-212 Bill works in the areas of computational biology and protein evolution. One of the founders of the field of sequence analysis and its applications for functional inference, he serves on many advisory boards, including that for European Bioinformatics Institute's UniProt, one of the two (with Genbank) primary public database for sequence information and annotation. He wrote the FASTA algorithm, one of the first widely used computer programs for identifying distantly related protein sequences. His approach is widely used to annotate newly sequenced genomes. His current research focuses on improving functional predictions for proteins by integrating annotation and variation information with sequence alignments. Title and abstract: Alignments with Annotations – providing functional information with similarity search Similarity searching with sequence alignment is the most reliable method for characterizing genomes and meta-genomes; sequence alignment algorithms may now be the major consumer of compute cycles in the life sciences. While similarity searching has become dramatically more sensitive over the past 25 years, and the number and diversity of sequences has increased by orders of magnitude, the display of sequence alignments is essentially unchanged. Despite the explosive growth of sequence, variation, and annotation data, and the hyper-connectedness of sequence and annotation databases, popular search programs show the same alignment information that was displayed in 1985. The FASTA search and alignment programs can incorporate functional site annotations, variation data, and domain boundaries Variant annotation allows the alignment score to improve; the domain annotation strategy partitions the alignment score across the annotated domains, reducing the problem of homologous over-extension. Annotation of functional sites allows exploration of the relationship between sequence similarity, active-site conservation, and functional similarity -- Patricia Babbitt Professor, Department of Bioengineering & Therapeutic Sciences and Department of Pharmaceutical Chemistry California Institute for Quantitative Biosciences Director, Graduate Program in Biological & Medical Informatics University of California, San Francisco UCSF MC 2550 Byers Hall Room508E 1700 4th Street San Francisco, CA 94158-2330, USA (CA 94158 for direct delivery by courier) Tel +1 (415) 476-3784 Fax +1 415-502-8193 Email: babbitt@cgl.ucsf.edu; Web http://babbittlab.ucsf.edu/ Assistant: Ms. Joslyn Polzien, +1 415-514-1214, 415-514-4861 Joslyn.Polzien@ucsf.edu