Genetic databases just might already have the answers youre seeking'

1
Genetic databases just might already have the
answers youre seeking.

• Minoli Perera Pharm.D, PhD
• Oct 6th, 2009

2
What can you do with just your computer!

• And possible a few text files.

3
Approaches in Association Studies

• SNP association
• Focused approached
• Uses candidate genes
• Mostly focused on functional SNPs
• N small to large.

Fromm MF et al., Pharmacogenetics and
Pharmacogenomics, 15 (10), 2005
Kim KA et al., Pharmacogenomics, 46 (1), 2007
4
(No Transcript)
5
(No Transcript)
6
PharmGKB
Central repository for genetic, genomic,
molecular and cellular phenotype data and
clinical information about people who have
participated in pharmacogenomics research studies.

• Publicly available internet research tool
• Funded by NIH
• Part of PGRN network
• Central repository of clinical and genetic
  information
• Limited to a few genes and pathways

• http//www.pharmgkb.org/index.jsp

7
rs Numbers

• Annotation used as a unique identifier of a SNP
  (usually the way it is reported in a paper).
• Can search dbSNP for information on particular
  SNPs.
• dbSNP is not completely reliabe as a source for
  SNP selection.

8
Seattle SNPs and other databases
http//pga.mbt.washington.edu/
9
SeattleSNPs Example
10
NIEHS SNP resequencing

• Several class of SNPs
• Resequenced in Coriell samples
• Many Pgx relevant genes

11
Haplotype Approach

• Haplotype association
• Most comprehensive
• Still used candidate gene
• N Large
• Usual restricted to single population

Reider MJ et al, NEJM, 352(22), 2005
12
Haplotypeswhat are they
A
T
G
G
C
A
A
T
G
A
G
T
A
C
G
G
T
C
A
T
G
G
T
A
13
Resources HapMap
14
Additional populations in Phase 3

• Population descriptorsASW (A) African ancestry
  in Southwest USACEU (C) Utah residents with
  Northern and Western European ancestry from the
  CEPH collectionCHB (H) Han Chinese in Beijing,
  ChinaCHD (D) Chinese in Metropolitan Denver,
  ColoradoGIH (G) Gujarati Indians in Houston,
  TexasJPT (J) Japanese in Tokyo, JapanLWK (L)
  Luhya in Webuye, KenyaMEX (M) Mexican ancestry
  in Los Angeles, CaliforniaMKK (K) Maasai in
  Kinyawa, KenyaTSI (T) Toscans in ItalyYRI (Y)
  Yoruba in Ibadan, Nigeria

15
Example
16
(No Transcript)
17
(No Transcript)
18
(No Transcript)
19

• As you click on each SNP, you can see what other
  SNP are binned with it.
• In this example 21 tag SNPs are needed to type 31
  SNPs from the original file.

20
HapMap

• Investigate known SNPs in any gene
• Download genotype data on any population
• Determine LD patterns and pick htSNPs
• Some genes not comprehensively surveyed

• http//www.hapmap.org/index.html
• SNP are lim ited and do not cover all regions of
  the gene.
• No resequencing data.

21
(No Transcript)
22
(No Transcript)
23
(No Transcript)
24
Genome Wide Approaches

• Genome Wide Association
• No a priori assumption on gene
• N Large to Very Large
• Usually a case/control
• SNP chips do not adequately cover all genes.

25
(No Transcript)
26
(No Transcript)
27
Other Databases

• Perlegen
• http//genome.perlegen.com/browser/index.html
• Genotype data, can also be found on HapMap
• dbSNP
• http//www.ncbi.nlm.nih.gov/projects/SNP/
• Can find rs for SNPs
• EntrezGene
• http//www.ncbi.nlm.nih.gov/entrez/query.fcgi?dbg
  enecmdsearchterm
• Can download reference sequences for genes on
  interest.

28
Considerations

• Extent of data What is well covered and what is
  missing
• What can you impute
• How can you use this to efficiently run your
  study
• Cheap GWAs

29
HapMap
Affy
Illum
DMET

• HapMap and Illumina fair equally well in finding
  variation in genes
• By extending the regions you look at in these
  platforms, you capture more of the SNPs found in
  sequence data.