introduction_to_gwas

Material for the Course “Introduction to genome-wide association studies (GWAS)”

Instructors: Filippo Biscarini, Oscar Gonzalez-Recio, Christian Werner

This course will introduce students, researchers and professionals to the steps needed to build an analysis pipeline for Genome-Wide Association Studies (GWAS). The course will describe all the necessary steps involved in a typical GWAS study, which will then be used to build a reusable and reproducible bioinformatics pipeline.

Each day the course will start at 14:00 and end at 20:00 (CET). As a general rule, we’ll have a longer break (30 minutes) at 16:00 and two shorter breaks (10-15 minutes) later on during the day (to be decided flexibly depending on the sessions).

Day 1

• Lecture 0 General Introduction / Overview of the Course [Filippo, Oscar, Christian]
  • General Introduction
  • GWAS Workflow (short)
• Lecture 1 GWAS Overview: Case Studies / Examples from Literature [Oscar]
  • GWAS Overview
• Lecture 2 Introduction to GWAS: Linkage Disequilibrium and Linear Regression [Oscar]
  • Introduction to GWAS
• Lab 1 (Demonstration) GWAS: Basic Models (Linear and Logistic Regression) [Oscar]
  • R code. Exercise on Simple Linear Regression
  • Rmarkdown Code. Exercise on Simple Logistic Regression
• Lab 2 - Tidyverse Introduction [Christian]
  • Tidyverse intro (pdf)
• Lab 3 - Description of Datasets [Christian]
  • Description of Datasets
• Course Manual
• GWAS Workflow

Day 2

• Lecture 3 The Multiple Testing Issue [Oscar]
  • Multiple Testing
  • R code. Exercise on multiple testing correction
• Lecture 4 Statistical Power, Population Stratification and Experimental Design [Oscar]
  • Statistical Power and Population Stratification
  • R code. Exercise on statistical power
• Lecture 5 Initial Data Analysis, Exploratory Data Analysis and Data Pre-Processing [Christian]
  • IDA, EDA & Data Pre-Processing–>
• Lab 3 GWAS: a first simple exercise for you! [Christian, Filippo]
  • [TBD]

Day 3

• Lab 4 Data filtering and mean/median imputation in R [Filippo]
  • [filter_genotype_data.R]
  • [mean_imputation.R]
  • [median_imputation.R]
• Lab 5 GWAS: The Stand-Alone Script(s) for the Full Model [Filippo]
  • [gwas_rrblup.R]
  • [gwas_statgengwas.R]
  • [gwas_sommer.R]
• Lecture 6 KNN Imputation
  • [KNN Imputation]
• Lab 6 (Demonstration) KNNI Imputation [Filippo]
  • [knni_illustration.Rmd]
  • [data_for_KNNI_illustration]
  • [knni_tidymodels.R]
  • [02_knni.sh] [support script]
  • [hamming.R] [support script]
  • [knni.R] [support script]
• Lecture 7 Data Types & Formats
  • [Common Data Types and Formats]
• Lab 7 EDA & IDA and preprocessing with Plink [Christian]
  • [Linux and the Shell]
  • [Unix Cheatsheet]
• Lecture 8 Imputation of Missing Genotypes [Christian]
  • [Imputation]
• Lab 8 Imputation of Missing Genotypes using Beagle [Christian]

Day 4

• Lecture 9 Brief Intermission:
  • [R code PCA & Population Structure]
  • [Imputed rice genotypes]
• Lab 9 Revising the Steps involved in GWAS [Filippo]
  • [slides]
  • [1.get_data.sh]
  • [2.step_filtering.sh]
  • [3.step_imputation.sh]
  • [4.gwas.sh]
• Lab 10 Introducing the Exercise [Filippo]
  • [Collaborative Exercise]
• Collaborative Exercise: let’s build our own GWAS workflow on new data [Filippo, Oscar, Christian]
  • Part 1: Individual/Group Break-Out Sessions to give it a try independetly
  • Part 2: Whole-Group Revision of the Exercise
• Lecture 10 Bioinformatics Pipelines: a super-elementary Introduction [Filippo]
  • [A bioinformatics pipeline for GWAS]
• Lab 11 Building a Pipeline with Snakemake [Filippo]
• Lab 12 The GWAS pipeline for Continuous Phenotypes [Filippo]
  • Plug-In for Mean or KNN Imputation
  • The GWAS pipeline for Binary Phenotypes (Guided Exercise) [Filippo]

Day 5

• Lecture 11 A light Touch on Post-GWAS Analysis: Inferring Functionality [Oscar]
  • [slides]
  • [R code. Exercise on R, and FUMA]
• Lecture 12 GWAS Model Extensions: [Filippo]
  • [12.1 GWAS Model Extensions_Dominance_and_other_genotype_Codifications]
  • [12.2 GWAS Model Extensions_Polyploids]
  • [12.3 GWAS Model Extensions_Trait_Types]
  • [12.4 GWAS Model Extensions_Multi-Trait-Locus, software]
  • [R code GWASpoly (vignette)]
  • [R code GWAS for categorical Traits]
  • [R code GWAS for categorical Traits - Examples]
  • [R code GWAS for longitudinal Traits]
  • [R code GWAS for multi-trait and multi-locus Models]

• Final Quiz on what we learned about GWAS! [Filippo, Oscar, Christian]
• Conclusions and Wrap-Up Discussion on GWAS [Filippo, Oscar, Christian]

Organization of the Code for the practical Sessions

1. the GWAS workflow in R
2. preparatory_steps: download and prepare the data
3. preprocessing: filter the data
4. imputation: imputing missing genotypes
5. gwas: run the GWAS models
6. power_and_significance: designing GWAS experiments
7. steps: identifying the individual steps involved in a GWAS study
8. pipeline: assembling the individual steps into a bioinformatics pipeline for GWAS
9. collaborative exercise: trying out what we learnt on new data