Genome-wide association mapping, spring 2009

Lecturer

Mikko Sillanpää

Scope

6-8 cu.

Type

Advanced studies.

Prerequisites

Basic concepts of probability calculus and of likelihood based methods in statistical inference. There are no molecular biological prerequisites for the course. Familiarity with the contents of the course "Statistical Methods in Genetics" is recommended, however.

Lectures

Weeks 3-9, Monday 10-12 in room C323, Thursday 10-12 in room B120.

Course folder, containing the course material other than the book, is available in room C326 of Exactum building.

Schedule

Monday 12.1.
 Introduction to the topic.
 What is the QTL?
How to build the genetic model to be used in association study?

Thursday 15.1.
 Multiple QTL model.
 Model selection.
Bayesian statistics.

Monday 19.1.
 "Classical allele-phenotype association testing"
 Collins-Book: Chapter 7: "Linkage disequilibrium mapping for
complex disease genes". pp.85-107.

Thursday 22.1.
 Correcting methods for population stratification I;
 (Structured association, Genomic control, TDT, matching, smoothing)

Monday 26.1.
 Correcting methods for population stratification II;
 -EM algorithm

Thursday 29.1.
 Mixed effect models and their use in correcting stratification
 and cryptic relatedness (see Yu et al. 2006).

• logit-model for discrete phenotypes
• LD-measures

Monday 2.2.
 Principles of Bayesian analysis and Markov Chain Monte Carlo.

Thursday 5.2.
 Visiting talk by Dr. Samuli Ripatti on the
topic: "Whose got the biggest N: meta-analysis of genome-wide association data across Europe and beyond".
 (in B120, 10.15-11.45) Students are encouraged to make lot of questions.

Monday 9.2.
 Haplotype-based association (see Zaykin et al. 2002) and
haplotype sharing methods (see pp. 290-291 in Thomas (2004))

Thursday 12.2.
 Replication in genetic studies of complex traits (see Sillanpää and Auranen 2004).

Monday 16.2.
 Case study in Chronique fatique syndrome (see Bhattacharjee et al. 2008).
 Imputation and prediction of pseudomarker genotypes at arbitrary map positions (see Marchini et al. 2007).

Thursday 19.2.
 Population-based haplotyping (phase and EM).
 Tag SNP-selection (see He and Zelikovsky 2006).
 Power of TDT vs population association (see Long and Langley 1999).
 Using allelic coding for finding epistatic interactions (see Sillanpää 2009).

Monday 23.2.
 Repetition of the some of the important topics in the course.

Thusday 26.2.
 Examn at B120 (Exactum), Note the time: 10.15-12.00!

Contents

The course covers statistical methods and issues involved in association mapping, with a special emphasis on genome-wide level genetic analyses. Replication (verification), confounding factors, and meta-analysis are considered as important themes in the course.

Exams

The examination will be at the last lecture Thursday, February 26th, from 10.15-12.00 in room B120 (Exactum).

Bibliography

Balding DA (2006) A tutorial on statistical methods for population association studies. Nature Reviews Genetics 7: 781-791.
 Bhattacharjee et al. (2008) Bayesian biomarker identification based on marker-expression proteomics data. Genomics 92: 384-392.
Yu et al. (2006) A unified mixed-model method for association mapping that accounts for multiple levels of relatedness. Nature Genetics 38: 203-208.
 Zaykin et al. (2002) Testing association in statistically inferred haplotypes with discrete and continuous traits in samples of
unrelated individuals. Human Heredity 53: 79-91.
 He J and Zelikovsky A (2006) MLR-tagging: informative SNP selection for unphased genotypes based on multiple linear regression.Bioinformatics 22: 2558-2561.
 Long AD and Langley CH (1999) The power of association studies to detect the contribution of candidate genetic loci to variation in complex traits. Genome Research 9: 720-731.
 Marchini et al. (2007) a new multipoint method for genome-wide association studies by imputation of genotypes. Nature Genetics 39: 906-913.
 Sillanpää MJ (2009) Detecting interactions in association studies by using simple allele recoding. Human Heredity 67: 69--75.
 Sillanpää MJ and Auranen K (2004) Replication in genetic studies of complex traits. Annals of Human Genetics 68: 646-657.
 Thomas DC (2004) Statistical Methods in Genetic Epidemiology, Oxford University Press.
 Collins AR (ed.)(2007): Linkage Disequilibrium and Association Mapping - analysis and applications, Humana Press.

Registration at the first lectures (January 12th-19th)

To complete the course

To score 6 credit points (6op), a student have to pass the examn and complete the practical work (essee) on some statistical topic(s) involved in the course. The length of the essee may be around 10-15 pages.

To score 8 credit points (8op), student have to pass the examn and complete the practical work (essee) on some statistical topic(s) involved in the course + perform (and document) an example genetic analysis with some real data set (students' own or public data). The example analysis can be done with some public genetic association analysis software (e.g., PLINK or TASSEL) or general statistical software like SAS, R or WinBUGS.

Evaluation (degree) of the course will be based on both the examn and the practical work.

The topics of the practical work (essee) may be consideration of 1-2 research papers, or some other statistical topic(s) involved in the course. It is OK if two/three students study the same topic together but eventually make their own practical works. Examples of possible topics of practical works based on research papers are for example:

1. "Favoring the detection of rare variants with high genetic effects in genome-wide association studies."

The paper describing the method:
 Dalmasso C, Genin E, Tregouet D-A (2008)
A weighted-Holm procedure accounting for allele frequencies in genomewide association studies.
 Genetics 180: 697-702.

2. "Combining evidence of natural selection with association analysis."

The paper describing the method:
 Ayodo G, Price AL, Keinan A, et al. (2007)
Combining evedence of natural selection with association analysis increases power to detect malaria-resistance variants.
 American Journal of Human Genetics 81: 234-242.

3. "Mixed effect models and their use in correcting stratification and cryptic relatedness in association studies."

The paper describing the method:
 Yu J, Pressoir G, Briggs W, et al. (2006)
A unified mixed-model method for association mapping that accounts for multiple levels of relatedness.
Nature Genetics 38: 203-208.