Data Mining in Bioinformatics Day 8: Feature Selection in - - PowerPoint PPT Presentation

▶

Sep 15, 2022 259 likes •431 views

Data Mining in Bioinformatics Day 8: Feature Selection in Bioinformatics Epistasis detection Karsten Borgwardt February 21 to March 4, 2011 Machine Learning & Computational Biology Group Max Planck Institutes Tbingen, Germany Karsten

SLIDE 1

Karsten Borgwardt: Data Mining in Bioinformatics, Page 1

Data Mining in Bioinformatics Day 8: Feature Selection in Bioinformatics

Epistasis detection

Karsten Borgwardt February 21 to March 4, 2011 Machine Learning & Computational Biology Group Max Planck Institutes Tübingen, Germany

SLIDE 2

Classic and new questions

Karsten Borgwardt: Data Mining in Bioinformatics, Page 2

Genetics How does genotypic vari- ation lead to phenotypic variation? Can we predict pheno- types based on the geno- type of an individual? Recent progress Genotypes can be de- termined at an unprece- dented level of detail Phenotypes can be recorded in an auto- mated manner

SLIDE 3

Phenotype prediction

Karsten Borgwardt: Data Mining in Bioinformatics, Page 3

Phenotype

AUCSVM

Chlorosis at 22◦C 0.629 ± 0.003 Anthocyanin at 16◦C 0.569 ± 0.003 Anthocyanin at 22◦C 0.609 ± 0.003 Leaf Roll at 10◦C 0.696 ± 0.002 Leaf Roll at 22◦C 0.587 ± 0.004

99-199 plants, 250k SNPs, Atwell et al., Nature 2010

Why is there room for improvement? We assume additive effects of SNPs, ignore gene-gene interactions. We ignore population structure, that is systematic an- cestry differences of cases and controls. We ignore gene-environment interactions.

SLIDE 4

Gene-gene interactions

Karsten Borgwardt: Data Mining in Bioinformatics, Page 4

Scale of the problem Typical datasets include order 105 − 106 SNPs. Hence we have to consider order 1010 − 1012 SNP pairs. Enormous multiple hypothesis testing problem. Enormous computational runtime problem. Our contribution We assume binary phenotypes (cases and controls). Genotypes may be homozygous or heterozygous. We assume m individuals with n SNPs each. We define an algorithm called epiSVM that rapidly de- tects epistatic interactions underlying the phenotypes in

O(mn).

SLIDE 5

Epistasis detection approaches

Karsten Borgwardt: Data Mining in Bioinformatics, Page 5

Filtering approaches Statistical criterion, e.g. SNPs with large main effect (Zhang et al., 2007) Biological criterion, e.g. underlying PPI (Emily et al., 2009) Index structure approaches fastANOVA, branch-and-bound on SNPs (Zhang et al., 2008) TEAM, efficient updates of contingency tables (Zhang et al., 2010) Exhaustive enumeration Only with special hardware such as GPU implementa- tions: EPIBLASTER (Kam-Thong, EJHG 2010)

SLIDE 6

Support Vector Machines

Karsten Borgwardt: Data Mining in Bioinformatics, Page 6

Separating hyperplane classifier (Vapnik & Chervonenkis, 1974)

Feature 1 Feature 2 w = (1,1) (x)

SLIDE 7

SVM: naive approach

Karsten Borgwardt: Data Mining in Bioinformatics, Page 7

Classify case/control by means of pairs of features SVM classifier:

f(x) = sgn(w, φ(x) + b) = sgn(

wγφγ(x) + b)

Mapping φ : X → X ⊗ X We have to compute all n2 entries of w to detect the feature pairs with maximum weight. These may be up to

1012 entries.

Can we avoid this?

SLIDE 8