Skip to content

Conversation

@dcdehaas
Copy link
Member

  1. Missing data support for non-standardized LinearOperators
  2. Standardized matrix (input) support for GWAS and GWAS with covariates
  3. Missing data support for GWAS (with and without covariates, with and without standardization)
  4. Other minor bug fixes related to CLI inputs/outputs

1. The standardized operators don't need any changes. Just pass in
   the missingness-adjusted allele frequencies when constructing
   the operators.
2. The non-standardized operators need an adjustmet before or after
   the multiplication, depending on the direction of the matmul().
   The user can pass in a mean value (e.g., the allele frequencies)
   when constructing the operator, and the operator will use it to
   perform the adjustments for the user.
1. Add linear regression-based GWAS with covariates method, and
   compare it to the QR method.
2. Adjust the beta computation for the QR method: it was missing
   the subtraction of the means, as per the derivation.
3. Ensure that you can run `grapp pca` and pipe the results
   directly in as covariate input to `grapp assoc`, by using
   Pandas-style tsv format if the covariates files ends in ".tsv"
   and plink-style format if the file ends in ".txt"
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

Successfully merging this pull request may close these issues.

2 participants