christacaggiano/treewas-exploration

exploration of TreeWas/TreeWas2 code for the Zaitlen lab. Based on TreeWas developed by McVean Lab @ Oxford.

Zaitlen Lab Exploration into TreeWas

UCSF

October 2018

---

Slides

• slides from my 10/4/18 journal club

TreeWas papers

This code draws from the methods described in the following papers:

• TreeWas: "Bayesian analysis of genetic association across tree-structured routine healthcare data in the UK Biobank" (Nat Gen, 2017)
• what I'm calling TreeWas2: "Systematic classification of shared components of genetic risk for common human diseases" (biorxiv, 2018)

Data

• Repository uses UKBiobank genotype and phenotype data. Phenotype data codes are 20002 (self-reported illness, non-cancer) and 41142 and 41078 (hospital episode statistics) Paths to local Zaitlenlab data:

• Disease classification comes from ICD10 codes

• Phenotype matrix of all diseases here: /ye/zaitlenlabstore/christacaggiano/treeWas-exploration/UKBIOBANK_DATA/

• Genotypes can be selected using this script for SNP(s) of interest: /ye/zaitlenlabstore/christacaggiano/treeWas-exploration/UKBIOBANK_DATA/scripts/generate_genotypes.sh

  • Parameters for this script: SNPs of interest (one on each line, ex: /ye/zaitlenlabstore/christacaggiano/treeWas-exploration/UKBIOBANK_DATA/snps.txt) and an output prefix.

  qsub generate_genotypes <file of SNPs of interest> <output prefix>
  
  # ex: qsub generate_genotypes snps.txt output_snps

  • afterward, filter to people of interest using scripts/select_people.py

     qsub select_people.sh <ids> <output>
  
     # ex: qsub select_people.sh ../british_ancestry_ids.txt ../icd10_phenotype_british.txt
  

• Age, sex, site of collection - british_covariates.txt

• Filtered matrix of just British Isles ancestry - icd10_phenotype_british.txt

• Filtered matrix of people who passed PLINK QC - icd10_passed_qc.txt

TreeWAS Scripts

• Original treeWAS code /ye/zaitlenlabstore/christacaggiano/treeWas-exploration/TreeWAS

• Install treewas with R -

  library(devtools)
  install_github("mcveanlab/TreeWAS/TreeWAS")

• Example TreeWAS code - /ye/zaitlenlabstore/christacaggiano/treeWas-exploration/treewas_example.R

• TreeWas2.0 code - /ye/zaitlenlabstore/christacaggiano/treeWas-exploration/TreeWas2.0

  library(devtools)
  install_github("christacaggiano/TreeWASDir")

• TreeWAS 2.0 Sample code-

• /ye/zaitlenlabstore/christacaggiano/treeWas-exploration/treewas2.0_example.R

• To save dynamic Plotly tree, click on export in Rstudio image window, and save as html.