You are on page 1of 42

Introduction to BLAST

David Fristrom Bibliographer/Librarian Science and Engineering Library 617 358-4124

What is BLAST?
Free, online service from National Center for Biotechnology Information (NCBI)

What is BLAST?


Nucleotide/Protein Sequence Databases

Google : Internet

Some Uses for BLAST

Identify an unknown sequence Build a homology tree for a protein Get clues about protein structure by finding similar proteins with known structures Map a sequence in a genome Etc., etc.

What is BLAST?

Basic Local Alignment Search Tool

AACGTTTCCAGTCCAAATAGCTAGGC ===--=== =-===-==-====== AACCGTTC TACAATTACCTAGGC Hits(+1): 18 Misses (-2): 5 Gaps (existence -2, extension -1): 1 Length: 3 Score = 18 * 1 + 5 * (-2) 2 2 = 6

Global Alignment
Compares total length of two sequences
Needleman, S.B. and Wunsch, C.D. A general method applicable to the search for similarities in the amino acid sequence of two proteins. J Mol Biol. 48(3):44353(1970).

Local Alignment
Compares segments of sequences Finds cases when one sequence is a part of another sequence, or they only match in parts.
Smith, T.F. and Waterman, M.S. Identification of common molecular subsequences. J Mol Biol. 147(1):195-7 (1981)

Search Tool
By aligning query sequence against all sequences in a database, alignment can be used to search database for similar sequences But alignment algorithms are slow

What is BLAST?
Quick, heuristic alignment algorithm Divides query sequence into short words, and initially only looks for (exact) matches of these words, then tries extending alignment. Much faster, but can miss some alignments
Altschul, S.F. et al. Basic local alignment search tool. J Mol Biol. 215(3):403-10(1990).

What is BLAST?
BLAST is not Google BLAST is like doing an experiment: to get good, meaningful results, you need to optimize the experimental conditions

Sample Search
Human beta globin (HBB)
Subunit of hemoglobin

Acquisition number: NP_000509 Limit to mouse to more easily show differences between searches

Interpreting Results
Score: Normalized score of alignment (substitution matrix and gap penalty). Can be compared across searches Max score: Score of single best aligned sequence Total score: Sum of scores of all aligned sequences

Interpreting Results
Query coverage: What percent of query sequence is aligned E Value: Number of matches with same score expected by chance. For low values, equal to p, the probability of a random alignment Typically, E < .05 is required to be considered significant

Getting the most out of BLAST

1. 2. 3. 4. What kind of BLAST? Pick an appropriate database Pick the right algorithm Choose parameters

Step 0: Do you need to use BLAST?

Step 1: Nucleotide BLAST vs. protein BLAST

Largely determined by your query sequence BUT If your nucleotide sequence can be translated to a peptide sequence, you probably want to do it (use tool such as ExPASy Translate Tool) Protein blasts are more sensitive and biologically significant

Sometimes it makes sense to use other blasts

Specialized Search: blastx

Search protein database using a translated nucleotide query Use to find homologous proteins to a nucleotide coding region Translates the query sequence in all six reading frames Often the first analysis performed with a newly determined nucleotide sequence

Specialized Search: tblastn

Search translated nucleotide database using a protein query Does six-frame translations of the nucleotide database Find homologous protein coding regions in unannotated nucleotide sequences such as expressed sequence tags (ESTs) and draft genome records (HTG)

Specialized Search: tblastx

Search translated nucleotide database using a translated nucleotide query Both translations use all six frames Useful in identifying potential proteins encoded by single pass read ESTs Good tool for identifying novel genes Computationally intensive

Even More Specialized

Make specific primers with Primer-BLAST Search trace archives Find conserved domains in your sequence (cds) Find sequences with similar conserved domain architecture (cdart) Search sequences that have gene expression profiles (GEO) Search immunoglobulins (IgBLAST) Search for SNPs (snp) Screen sequence for vector contamination (vecscreen) Align two (or more) sequences using BLAST (bl2seq) Search protein or nucleotide targets in PubChem BioAssay Search SRA transcript libraries Constraint Based Protein Multiple Alignment Tool

Step 2: Choose a Database

Too large:
Takes longer Too many results More random, meaningless matches

Too small or wrong one:

Miss significant matches

Protein Databases
Non-redundant protein sequences (nr)
Translations of GenBank coding sequences (CDS) RefSeq Proteins PDB (RCSB Protein Data Bank - 3d-structure) SwissProt Protein Information Resource (PIR) Protein Research Foundation (Japanese DB)

Reference proteins (refseq_protein)

NCBI Reference Sequences: Comprehensive, integrated, nonredundant, well-annotated set of sequences

Swissprot protein sequences (swissprot)

Swiss-Prot: European protein database (no incremental updates)

Protein Databases
Patented protein sequences (pat)
Patented sequences

Protein Data Bank proteins (pdb)

Sequences from RCSB Protein Data Bank with experimentally determined structures

Environmental samples (env_nr)

Protein sequences from environmental samples (not associated with known organism)

Nucleotide Databases
Human genomic + transcript

Mouse genomic + transcript

Nucleotide collection (nr/nt)

nr stands for non-redundant, but it isnt
GenBank (NCBI) EMBL (European Nucleotide Sequence Database) DDBJ (DNA Databank of Japan) PDB (RCSB Protein Data Bank - 3d-structure)

Kitchen sink but not HTGS0,1,2, EST, GSS, STS, PAT, WGS

Nucleotide Databases
Reference mRNA sequences (refseq_rna) Reference genomic sequences (refseq_genomic)
NCBI Reference Sequences: Comprehensive, integrated, non-redundant, well-annotated set of sequences

NCBI Genomes (chromosome)

Complete genomes and chromosomes from Reference Sequences

Nucleotide Databases
Expressed sequence tags (est) Non-human, non-mouse ESTs (est_others)

Genomic survey sequences (gss)

Like EST, but genomic rather than cDNA (mRNA)
random "single pass read" genome survey sequences. cosmid/BAC/YAC end sequences exon trapped genomic sequences Alu PCR sequences transposon-tagged sequences

Nucleotide Databases
High throughput genomic sequences (HTGS)
Unfinished sequences (phase 1-2). Finished are already in nr/nt

Patent sequences (pat)

Patented genes

Protein Data Bank (pdb)

Sequences from RCSB Protein Data Bank with experimentally determined structures

Nucleotide Databases
Human ALU repeat elements (alu_repeats)
Database of repetitive elements

Sequence tagged sites (dbsts)

Short sequences with known locations from GenBank, EMBL, DDBJ

Whole-genome shotgun reads (wgs)


Nucleotide Databases
Environmental samples (env_nt)
Nucleotide sequences from environmental samples (not associated with known organism)

Database Options
Limit to (or exclude) an organism Exclude Models (XM/XP)
Model reference sequences produced by NCBI's Genome Annotation project. These records represent the transcripts and proteins that are annotated on the NCBI Contigs which may have been generated from incomplete data.

Entrez Query
Use Entrez query syntax to limit search

Step 3: Choose an Algorithm

How close a match are you looking for? Determines how similarities are scored Affects speed of search and chance of missing match Again, what is the goal of the search?

Protein-protein BLAST Standard protein BLAST

Protein-protein BLAST Position-Specific Iterated BLAST Finds more distantly related matches Iterates: Initial search results provide information on allowed mutations; subsequent searches use these to create custom substitution matrix

Protein-protein BLAST Pattern Hit Initiated BLAST Variation of PSI-BLAST Specify a pattern that hits must match Use when you know protein family has a signature pattern: active site, structural domain, etc. Better chance of eliminating false positives Example: VKAHGKKV

Nucleotide BLAST Finds highly similar sequences Very fast Use to identify a nucleotide sequence

Nucleotide BLAST Use to find less similar sequences

discontiguous megablast
Nucleotide BLAST
Bioinformatics. 2002 Mar;18(3):440-5. PatternHunter: faster and more sensitive homology search. Ma B, Tromp J, Li M.

Even more dissimilar sequences Use to find diverged sequences (possible homologies) from different organisms

Step: 4 Algorithm Parameters

Fine-tune the algorithm Short Queries Expect threshold: The lower it is, the fewer false positives (but you might miss real hits)

Algorithm Parameters
Scoring Matrix: PAM: Accepted Point Mutation
Empirically derived chance a substitution will be accepted, based on closely related proteins Higher PAM numbers correspond to greater evolutionary distance

BLOSUM: Blacks Substitution Matrix

Another empirically derived matrix, based on more distantly related proteins Lower BLOSUM numbers correspond to greater evolutionary distance

Compositional adjustment changes matrix to take into account overall composition of sequence

Algorithm Parameters
Filters and Masking Can ignore low complexity regions in searching

Additional Sources

Pevsner, Jonathan Bioinformatics and Functional Genomics, 2nd ed. (Wiley-Blackwell, 2009) BLAST help pages: http:// Web&PAGE_TYPE=BlastDocs Slides from class on similarity searching; lots of technical details on algorithms and similarity matrices: