生物信息学国外教程2010.pptVIP

* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * BLAT Fig. 5.20 Page 167 LAGAN Fig. 5.21 Page 168 SSAHA Outline of today’s lecture BLAST Practical use Algorithm Strategies Finding distantly related proteins: PSI-BLAST Hidden Markov models BLAST-like tools for genomic DNA PatternHunter Megablast BLAT, BLASTZ Where we are in the course --We started with “access to information” (Chapter 2) --We next covered pairwise alignment (Chapter 3), then BLAST in which one sequence is compared to a database (Chapters 4, 5) --Next we’ll describe multiple sequence alignment (Chapter 6) --We’ll then visualize multiple sequence alignments as phylogenetic trees (Chapter 7). That topic spans molecular evolution. * * * * * * * * * * * * * * * * * * * * * * * * * * * Scoring matrices let you focus on the big (or small) picture retinol-binding protein retinol-binding protein PAM250 PAM30 Blosum45 Blosum80 Fig. 5.7 Page 151 PSI-BLAST generates scoring matrices more powerful than PAM or BLOSUM retinol-binding protein retinol-binding protein Fig. 5.7 Page 151 PSI-BLAST: performance assessment Evaluate PSI-BLAST results using a database in which protein structures have been solved and all proteins in a group share 40% amino acid identity. Page 150 PSI-BLAST: the problem of corruption PSI-BLAST is useful to detect weak but biologically meaningful relationships between proteins. The main source of false positives is the spurious amplification of sequences not related to the query. For instance, a query with a coiled-coil motif may detect thousands of other proteins with this motif that are not homologous. Once even a single spurious protein is included in a PSI-BLAST search above threshold, it will not go away. Page 152 PSI-BLAST: the problem of corruption Corruption is defined as the presence of at least one false positive alignment with an E value 10-4 after five iterations. Three approaches to