A concise description of why we're pursuing this task.
Dave Cutler's recommendations for stuff they would find meaningful, and which we should do our best to get in before end of January.
Completion Criteria:
What is the goal, and how do we know this task is accomplished?
[ ] Add pLi and Loeuf scores for genes.
[ ] Add MPC scores for variants.
[ ] Allow the user to upload a pedigree file with their VCF, and then output the plink2 files directly, with the correct pedigree /
sex / phenotype information, and the correct "names" (rs numbers when available,
and something natural like chr_position for variants with no rs number) for the
variants.
[ ] Add "SKAT" input to the plink data. Basically let the user pick variant criteria, and
output lists of those variants broken down by gene... something like
that.... not redumping the whole VCF or annotations or plink files, but literally
just variant lists by gene, readable by SKAT, where the user specified the criteria
to make those lists.
[ ] For ancestry stuff you should think about doing your graphical wizardry.
How about having a set of 1000 genome samples precomputed and including
PC1 vs PC2, PC1 vs PC3, and PC2 vs PC3 plots with all the samples included
in the current VCF analysis... with current samples in a prominent color, and the public
samples labeled in a sort of light transparent-ish kinda color? Can I click on
points, and think about "flagging / excluding" them? I could actually imagine
replacing your tile plots on the front page with these PCA type plots.
If the pedigree file had some standardized field for self-reported ancestry, you
could even flag discrepancies between self-report and the apparent genetics.
[ ] Along the lines of ancestry stuff, you could add checks for duplicates / close
relatives, a sex check, etc, but Plink2 will do this plenty well, as is.
akotlar
commented
11 months ago
Background:
A concise description of why we're pursuing this task.
Dave Cutler's recommendations for stuff they would find meaningful, and which we should do our best to get in before end of January.
Completion Criteria:
What is the goal, and how do we know this task is accomplished?
[ ] Add pLi and Loeuf scores for genes.
[ ] Add MPC scores for variants.
[ ] Allow the user to upload a pedigree file with their VCF, and then output the plink2 files directly, with the correct pedigree / sex / phenotype information, and the correct "names" (rs numbers when available, and something natural like chr_position for variants with no rs number) for the variants.
[ ] Add "SKAT" input to the plink data. Basically let the user pick variant criteria, and output lists of those variants broken down by gene... something like that.... not redumping the whole VCF or annotations or plink files, but literally just variant lists by gene, readable by SKAT, where the user specified the criteria to make those lists.
[ ] For ancestry stuff you should think about doing your graphical wizardry. How about having a set of 1000 genome samples precomputed and including PC1 vs PC2, PC1 vs PC3, and PC2 vs PC3 plots with all the samples included in the current VCF analysis... with current samples in a prominent color, and the public samples labeled in a sort of light transparent-ish kinda color? Can I click on points, and think about "flagging / excluding" them? I could actually imagine replacing your tile plots on the front page with these PCA type plots. If the pedigree file had some standardized field for self-reported ancestry, you could even flag discrepancies between self-report and the apparent genetics.
[ ] Along the lines of ancestry stuff, you could add checks for duplicates / close relatives, a sex check, etc, but Plink2 will do this plenty well, as is.