Transcript Medical Applications of Bioinformatics

Bioinformatics Tools
Stuart M. Brown, Ph.D
Dept of Cell Biology
NYU School of Medicine
Bioinformatics Tools
Stuart M. Brown, Ph.D
Dept of Cell Biology
NYU School of Medicine
Overview
This lecture will summarize a huge amount of
bioinformatics material that is usually presented as
a full 12 week course.
– Data management and analysis of sequences
from the HGP
– A quick look at GenBank and ENTREZ.
– Gene finding and translation
– Similarity searching and alignment (BLAST)
– Protein structure and function
Data Management and Analysis
• The Human Genome Project has generated
huge quantities of DNA sequence data.
• This data will lead to many medial advances.
• But a great deal of analysis and research will
be needed.
Access to the Data
•Organize the genome data & provide
access for scientists
•Use the Internet
• The data is public, so anyone can
access it.
GenBank
•All Genome Project data is stored in a database called
GenBank managed by the National Center for
Biotechnology Information (NCBI)
•The NCBI is a branch of the National Library of Medicine,
which is part of the NIH (National Institutes of Health).
http://ncbi.nlm.nih.gov
GenBank Sections
In addition to DNA sequences of genes
GenBank has a number of other sections
including:
• Protein sequences (translated from DNA)
• Short RNA fragments (ESTs)
• Cancer Genome Anatomy Project (CGAP) gene
expression profiles of normal, pre-cancer, and cancer
cells from a wide variety of tissue types
• Single Nucleotide Polymorphisms (SNPs) which
represent genetic variations in the human population
• Online Mendelian Inheritance in Man (OMIM) a
database of human genetic disorders
Finding Genes
•GenBank contains approximately 13 billion
bases in 12 million sequence records (as of
August 2001).
•These billions of G, A, T, and C letters would
be almost useless without descriptions of
what genes they contain, the organisms they
come from, etc.
•All of this information is contained in the
"annotation" part of each sequence record.
Entrez is a Tool for Finding
Sequences
• NCBI has created a Web-based tool called
Entrez for finding sequences in GenBank.
• Each sequence in GenBank has a unique
“accession number”.
• Entrez can also search for keywords such as
gene names, protein names, and the names
of orgainisms or biological functions
Entrez has links to Medline
• Entrez is much more than just a tool for
finding sequences by keywords.
• It contains links to PubMed/Medline
• Entrez also contains all known protein
sequences and 3-D protein structures.
Entrez is Internally Cross-linked
• DNA and protein sequences are linked
to other similar sequences
• Medline citations are linked to other
citations that contain similar keywords
• 3-D structures are linked to similar
structures
• These relationships might include genes in a
multi-gene family, related journal articles, or
other proteins in the same biochemical
pathway
• This potential for horizontal movement through
the linked databases makes Entrez a dynamic
tool.
• You can start with only a vague set of
keywords or a sequence from the laboratory
and rapidly access a set of relevant literature
and related database sequences.
Similarity Searching
• There are a variety of computer
programs that are used for making
comparisons between DNA sequences.
• The most popular is known as BLAST
(Basic Local Alignment Search Tool)
• BLAST is free at the NCBI website
BLAST Searches GenBank
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 is Complex
• Similarity searching relies on the concepts of
alignment and distance between pairs of
sequences.
• Distances can only be measured between
aligned sequences (match vs. mismatch at
each position).
• A similarity search is a process of testing the
best alignment of a query sequence with
every sequence in a database.
Search with Protein not DNA
1) 4 DNA bases vs. 20 amino acids - less
random similarity
2) Can have varying degrees of similarity
between different AAs
- # of mutations, chemical similarity, PAM
matrix
3) Protein databanks are much smaller
than DNA databanks.
BLAST has Automatic
Translation
• BLASTX makes automatic translation (in all
6 reading frames) of your DNA query
sequence to compare with protein databanks
• TBLASTN makes automatic translation of an
entire DNA database to compare with your
protein query sequence
• Only make a DNA-DNA search if you are
working with a sequence that does not code
for protein.
•
•
>gb|BE588357.1|BE588357 194087 BARC 5BOV Bos taurus cDNA 5'.
Length = 369
272 bits (137),
Expect = 4e-71
•
Score =
•
•
Identities = 258/297 (86%), Gaps = 1/297 (0%)
Strand = Plus / Plus
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
Query: 17
Sbjct: 1
Query: 77
Sbjct: 60
aggatccaacgtcgctccagctgctcttgacgactccacagataccccgaagccatggca 76
|||||||||||||||| | ||| | ||| || ||| | |||| ||||| |||||||||
aggatccaacgtcgctgcggctacccttaaccact-cgcagaccccccgcagccatggcc 59
agcaagggcttgcaggacctgaagcaacaggtggaggggaccgcccaggaagccgtgtca 136
|||||||||||||||||||||||| | || ||||||||| | ||||||||||| ||| ||
agcaagggcttgcaggacctgaagaagcaagtggagggggcggcccaggaagcggtgaca 119
Query: 137 gcggccggagcggcagctcagcaagtggtggaccaggccacagaggcggggcagaaagcc 196
|||||||| | || | ||||||||||||||| ||||||||||| || ||||||||||||
Sbjct: 120 tcggccggaacagcggttcagcaagtggtggatcaggccacagaagcagggcagaaagcc 179
Query: 197 atggaccagctggccaagaccacccaggaaaccatcgacaagactgctaaccaggcctct 256
||||||||| | |||||||| |||||||||||||||||| ||||||||||||||||||||
Sbjct: 180 atggaccaggttgccaagactacccaggaaaccatcgaccagactgctaaccaggcctct 239
Query: 257 gacaccttctctgggattgggaaaaaattcggcctcctgaaatgacagcagggagac 313
|| || ||||| || ||||||||||| | |||||||||||||||||| ||||||||
Sbjct: 240 gagactttctcgggttttgggaaaaaacttggcctcctgaaatgacagaagggagac 296
Understand the Statistics!
• BLAST produces an E-value for every match
– This is the same as the P value in a statistical test
• A match is generally considered significant if the
E-value < 0.05 (smaller numbers are more significant)
• Very low E-values (e-100) are homologs or
identical genes
• Moderate E-values are related genes
• Long regions of moderate similarity are more
important than short regions of high identity.
BLAST is Approximate
• BLAST makes similarity searches very
quickly because it takes shortcuts.
– looks for short, nearly identical “words” (11 bases)
• It also makes errors
– misses some important similarities
– makes many incorrect matches
• easily fooled by repeats or skewed composition
Bad Genome Annotation
• Gene finding is at best only 90% accurate.
• New sequences are automatically annotated
with BLAST scores.
• Bad annotations propagate
• Its going to take us 10-20 years or more to
sort this mess out!
Protein Function
• The ultimate goal of the HGP is to identify all
of the genes and determine their functions
• Genes function by being translated into
proteins:
– structural
– enzymes
– regulatory
– signalling
Translation
• Once we have found the DNA sequence of a gene,
we can decode the amino acid sequence of the
corresponding protein .
• The “Genetic Code” is actually quite simple.
Chemical Properties
Some chemical properties of a protein can be
calculated from its amino acid sequence:
• molecular weight
• charge/pH
• hydrophobicity
Patterns in Proteins
Conserved Domains
• Proteins are built out of functional units
know as domains (or motifs)
• These domains have conserved sequences
Often much more similar than their respective proteins
 Exon splicing theory (W. Gilbert)
• Exons correspond to folding domains which in turn
serve as functional units
• Unrelated proteins may share a single similar exon
(i.e.. ATPase or DNA binding function)

Simple Structures
Some motifs form structures that can be
recognized as simple sequence patterns:
–transmembrane domains
–coiled coils
–helix-turn-helix
–signal peptides
Functional Motifs
• Other functional portions of proteins can be
recognized by their sequence, even if their 3D structure is not known.
• There are many databases of protein
motifs/domains: ProSite, Pfam, ProDom, etc.
Tools for Finding Motifs
• Define a motif from a set of known proteins
that share a similar sequence and function.
• A pattern is a list of amino acids that can
occur at each position in the motif.
• A profile is a matrix that assigns a value to
every amino acid at every position in the
motif.
• A HMM is a more complex profile based on
pairs of amino acids.
Protein 3-D Structure
Structure = Function
• Proteins function by 3-D interactions with
other molecules (i.e. physical chemistry).
• So for a protein, 3-D structure is function.
• But we can’t accurately determine 3-D
structure from gene sequence.
Structure Prediction
Predicting a protein’s 3-D structure from its
amino acid sequence is incredibly complex.
– proteins are polypeptides (long chains of amino
acids)
– can fold and rotate around bonds within each
amino acid as well as the bonds between them
– it is not possible to evaluate every possible folding
pattern for an amino acid sequence
Secondary Structure
• The local structure of the amino acids in a
protein can also be predicted to some
extent.
• Each amino acid has a tendency to form
either an alpha helix or a beta sheet
detail:
....,....1....,....2....,....3....,....4....,.
AA
|MMSGAPSATQPATAETQHIADQVRSQLEEKYNKKFPVFKAVSFKSQ
PHD sec |
HHHHHHHHHHHHHHHH
EEEEEEEEEE
Rel sec |9999978996675999999999899976558778433688899999
prH
prE
prL
subset: SUB
sec
sec
sec
sec
|0000000002212899999999899987620111110000000000
|0000000000000000000000000000100000235788899898
|9998988897776000000000100011268888653111100000
|LLLLLLLLLLLLLHHHHHHHHHHHHHHHHLLLLL...EEEEEEEEE
ACCESSIBILITY
3st:
P_3 acc
10st:
PHD acc
Rel acc
subset: SUB acc
|bbebbeeeeeebbeebbebbeebeeebeeeeeee eebebbebebb
|0070067786700770070077067607777777377070070600
|1030213432520446046446724245555476154444252121
|.......e..e..eeb.ebbeeb.e.beeeeeee.eebeb.e....
Threading
• Rather than computing a 3-D structure from
scratch, it may be possible to find a similar
structure.
• Must have ~25% aa sequence identity.
• Uses a process called threading to create a
new structure based on a known structure.
• This still requires HUGE amounts of
computer power.
Protein Data Base
• There is a database of all known protein
structures called the PDB.
• These have been determined by X-ray
crystalography and/or NMR.
• Anyone download and view these structures
with a PDB viewer program.
RasMol
RasMol is the simplest PDB viewer.
http://www.umass.edu/microbio/rasmol/
It can work together with a web browser to let you
view the structure of any sequence found with
Entrez that has a known 3-D structure.
Gene Finding & Translation
• How can we find genes on
chromosomes?
• Genome project data is just huge
chunks of DNA.
• Does automatic annotation work?
Raw Genome Data:
Finding Genes is Not Easy
• Perhaps 1% of human DNA encodes
functional genes.
• Genes are interspersed among long
stretches of non-coding DNA.
• Repeats, pseudo-genes, and introns
confound matters
Pattern Finding Tools
It is possible to use DNA sequence patterns
to predict genes:
• Promoters
• translational start and stop codes (ORFs)
• intron splice sites
• codon usage
Similarity to Known Genes
• It is also possible to scan new DNA
sequence for known genes
• Can look for annotated genes/proteins
• Or just for RNAs (ESTs)