Applying haplotype models to association study design

1
Applying haplotype models to association study
design

• Natalie Castellana
• June 7, 2005

2
Background

• Certain characteristics are linked to genetic
  factors.

By finding models for genetic variation, we can
determine which genes render an individual
susceptible to a certain disease.
3
Background(2)
1000011010110100000010
Given that this sample has haplotype 01101, does
it have the disease?
4
Genetic Variation
Mutation
Recombination
1000011
1110101
Because of recombination, similar genetic
variation can be found within closely linked
regions.
5
Generating Data

• Generate genetic segments
• Isolate the disease causing allele, and segregate
  the case (diseased) samples from the control
  (healthy) samples.

1 0 1 1 1 0 1 1 1 1
Control
Case
1 1 1 0 0 1 1 0 1 0
0 0 1 1 0 0 1 1 0 1
1 0 1 0 1 1 0 0 1 0
6
Testing individual SNPs

• Go through each SNP and determine which SNPs
  accurately predict which samples have the disease
  and which do not.
• Case Control
• 10 11.. 010 0
• 01 10...... 100 1

7
Haplotype block method

• Instead of looking at each individual SNP, we can
  look at groups of contiguous SNPs.
• 110100000011
• 110110010001
• 011100000010
• 110110010000

8
Blocks vs. SNPs
9
Haplotype motif method

• Notion that a sequence is the concatenation of
  segments (like the block method) but does not
  require conservation of boundaries.
• 1101000000
• 1100100100
• 0111000000
• 1101100111

10
Approximation Algorithm

• General idea

Pick the best partition, minimizing the number of
motifs needed to explain all the data.
11
Finding Motifs
0 0 0 1 1 0 0 1
0 1 1 0 1 0 0 1 1
12
What I am working on

• Implementing an efficient algorithm for finding
  motifs.
• Performing association tests on the SNP, block,
  and motif models.