This Lecture
       Introduction to Biostatistics and Bioinformatics
           Sequence Database Searching
            Stuart M. Brown, Ph.D.
         Center for Health Informatics and Bioinformatics
              NYU School of Medicine
                 BLAST Searches GenBank
                    (or a custom database)
    [BLAST= Basic Local Alignment Search Tool]
    The NCBI BLAST web server lets you compare 
      your query sequence to various sections of 
      GenBank:
            –nr = non-redundant (main sections)
            –month = new sequences from the past few weeks
            –ESTs
            –human, drososphila, yeast, or E.coli genomes
            –proteins (by automatic translation)
    • This is a VERY fast and powerful computer.
             BLAST
  • Uses word matching
  • Similarity matching of words (3 aa’s, 11 bases) 
    – does not require identical words.
  • If no words are similar, then no alignment
    – won’t find matches for very short sequences 
  •  “gapped BLAST”  (BLAST 2) improved handling of gaps 
   in alignment
  • BLAST searches can be sent to the NCBI’s server from 
   website, or a custom client program (Unix)
      BLAST Algorithm
               BLAST Word Matching
       MEAAVKEEISVEDEAVDKNI
       MEA
        EAA       Break query
         AAV
          AVK      into words:
           VKE
            KEE
             EEI
              EIS
               ISV        Break database 
            ...
                            sequences
                            into words: