search

statgen / pheweb

A tool to build a website to browse hundreds or thousands of GWAS.
MIT License
154 stars 65 forks source link

best practice for adding new summary statistics files #174

Closed Shicheng-Guo closed 2 years ago

Shicheng-Guo commented 2 years ago

I have deposited about 2000 phenotypes into my personal pheweb server few month ago. However, now I need to deposit another 200 summary statistics files. Is there any suggestion how to do this task with high efficiency.

Thanks Shicheng

pjvandehaar commented 2 years ago

No, sorry. Just update pheno-list.json and re-run pheweb process and it'll regenerate what you need.

If you want to leave the original site running while you load the new one, you can copy it with hardlinks using cp -l -r. I don't think that anything in pheweb overwrites files in-place, but if it does then that will cause problems.

The parsing (pheweb parse which populates generated-by-pheweb/parsed/*) won't need to happen again, but everything else will.

PheWeb could have been written to make adding phenotypes fast, but it wasn't. It would need (1) a fast rsid annotator using <2GB RAM and (2) a way to update the big pheno-vs-variant matrix that's used for single-variant PheWAS lookups.

Shicheng-Guo commented 2 years ago

Clear!! Thanks @pjvandehaar