• Choices
  • IDPs
  • Phenotypes
    • Risk factors
    • Diseases
• Data used
  • IDPs
  • Phenotypes
• Project structure

Choices

IDPs

• White Matter Hyperintensity (WMH):
  • total volume
  • lesion count (only available in older stats with a smaller sample)
• QSM and T2* for WMH
• QSM for (left and right):
  • Caudate
  • Putamen
  • Pallidum
  • Substantia Nigra
  • Amygdala

Phenotypes

Risk factors

• 6150_4 hypertension (diagnosed)
• Diastolic blood pressure, automated reading (4079_irnt)
• Systolic blood pressure, automated reading (4080_irnt)
• Glucose (mmol/L) (30740_irnt)
• Diabetes diagnosed by doctor (2443)
• HDL cholesterol (mmol/L) (30760_irnt)
• LDL direct (mmol/L) (30780_irnt)
• Triglycerides (mmol/L) (30870_irnt)
• Body mass index (BMI) (21001_irnt)
• Filtered coffee intake (100270)
• Current tobacco smoking (1239)
• Sleep duration (1160)
• Alcohol intake frequency (1558)
• Time spent doing moderate physical activity (104910)

Diseases

• Vascular/heart problems diagnosed by doctor: Stroke (6150_3)

Data used

IDPs

• BIG40: https://open.win.ox.ac.uk/ukbiobank/big40/

  • WMH total volume:
    • IDP_T2_FLAIR_BIANCA_WMH_volume
    • file 1437

• All other IDPs from: https://www.fmrib.ox.ac.uk/ukbiobank/gwas_resources/index.html

Phenotypes

https://docs.google.com/spreadsheets/d/1kvPoupSzsSFBNSztMzl04xMoSC3Kcx3CrjVf4yBmESU/edit#gid=227859291

Project structure

1. QSM and T2* IDPs are downloaded in two separate folders: discovery and replication. The files from both samples are merged into one file per IDP via general/merge_samples.R.

2. Pasting variants.tsv to phenotypes summary stats via general/merge_variants.sh.

3. Various files need to be reformatted/polished/modified before running the other scripts. So run general/sanitize_phen.py before continuing.

4. LDSC

   • ldsc/munge.sh
   • ldsc/ld_score_regression.sh

5. Run ldsc/log_parser.py to generate a summary.txt file for the next step.

6. Visualize results of LDSC with ldsc/heatmap.py.

7. PascalX

   • Download reference panel by running *pascalx/download_ref.sh"
   • Download genome annotation by running pascalx/download_annotation.py
   • Xscoring with pascalx/xscorer.py