Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

If the PRS deciles still put the highest prevalence of ischemic stroke in the lowest decile group - check the following #9

Open
jackosullivanoxford opened this issue Apr 30, 2019 · 14 comments

Comments

@jackosullivanoxford
Copy link
Collaborator

  1. How did we create the phenotype file? Did the outcome (ischemic stroke) get flipped to the opposite?
@jackosullivanoxford
Copy link
Collaborator Author

jackosullivanoxford commented Jun 17, 2019

  1. Check the alleles in the GWAS summary statistics
  2. Check I have checked for ambiguous strands (A/T or C/G) in GWAS summary statistics, e.g. from Khera: "DNA polymorphisms with ambiguous strands (A/T or C/G) were removed from the score derivation." ?Talk to Anna/others about excluding SNPs where the effect allele does not match between GWAS SS and LD reference genotype file and validation genotype file

@jackosullivanoxford
Copy link
Collaborator Author

  1. Consider using PLINK2 to sum the betas across all the variants (instead of step3 in ldpred)

@jackosullivanoxford
Copy link
Collaborator Author

jackosullivanoxford commented Jul 3, 2019

  1. In step 1, when we tried to get the chromosome and position for the GWAS SS SNPs, we merged with a random 2000 sample of UKBB bim file. We lost a number of SNPs ~700,000. Try and merge these without losing SNPs from GWAS SS file. The raw plink files appear to be in here (but they are divided by chromosome): /oak/stanford/groups/euan/projects/ukbb/data/genetic_data/v2/plink
  • I have now done this: see R script ~/Documents/Stanford work/AF_G/2nd_run_through/GWAS_SS_CES.R

@jackosullivanoxford
Copy link
Collaborator Author

  1. In step 1, instead of using 1-Freq1 for reffrq, use Freq1. Try this on the AIS run through and see if it changes things!

@jackosullivanoxford
Copy link
Collaborator Author

  1. Check how I have created the deciles

@jackosullivanoxford
Copy link
Collaborator Author

jackosullivanoxford commented Jul 17, 2019

  1. Exclude warfarin patients: do ICD codes for warfarin and self-reported wafarin taking: https://biobank.ctsu.ox.ac.uk/crystal/coding.cgi?id=4&nl=1 (CONsider making ICD and self-reported separate columns, for medication but also CHADVASc)

@jackosullivanoxford
Copy link
Collaborator Author

jackosullivanoxford commented Jul 17, 2019

  1. Add in self-reported diagnoses of AF and ischemic stroke. This is the UKBB data showcase for self-reported codes: http://biobank.ctsu.ox.ac.uk/crystal/coding.cgi?id=6

October 2019: THIS IS NOW DONE

@jackosullivanoxford
Copy link
Collaborator Author

  1. Still need to divide strokes into incident and prevalent cases.

@jackosullivanoxford
Copy link
Collaborator Author

  1. Check HLA regions are excluded from AF genotypes (step 3 of ldpred)

@jackosullivanoxford
Copy link
Collaborator Author

  1. Consdier adding other event outcomes (e.g. PE, claudication etc)

@jackosullivanoxford
Copy link
Collaborator Author

  1. Consider removing self-reported and ICD TIA (and possibly generic self-reported 'stroke') from phenotype file.

@jackosullivanoxford
Copy link
Collaborator Author

  1. Add in ICD9 codes for a) AF, b) stroke (outcome) and c) CHADSVASc codes

@jackosullivanoxford
Copy link
Collaborator Author

jackosullivanoxford commented Oct 1, 2019

  1. Consider adding in self-reported hypertension meds and also BP readings (with adjustment): See 'PRS_stroke.pdf': Also include patients that self-reported taking anti-HTN meds as per meta-stroke (PRS_stroke.pdf): For hypertension we used an expanded definition including self-reported high blood pressure (either on blood pressure medication, data fields #6177, #6153; or systolic blood pressure >140 mmHg, fields #4080, #93; or diastolic blood pressure >90 mmHg, data fields #4079, #94) as well as hospital records # I think an argument against including BP as a continuous variable is the fact that CHADSVASc considers BP as a binary variable

@jackosullivanoxford
Copy link
Collaborator Author

  1. ADD algorithmically defined stroke outcomes to phenotype file and then re-run step 3

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant