Periodically generating PCoAs of interest for the stats page

[wasade]

It would be really cool if Qiita constructed new database-wide PCoAs on a nightly basis, and exposed the resulting .qzvs on the stats page.

There are a few perspectives that may be nice:

• all public
• all public/private
• specific studies like Microsetta

For display of private data, it will be necessary to anonymize the study and sample IDs.

Storing the resulting PCoAs overtime will allow for easily producing derived figures, such as the convex hull volume growth.

GPU UniFrac is fast enough to be able to do this compute. And it recently was modified to enable a PCoA only output, which is valuable here as we don't need to save the resulting distance matrix.

[justinshaffer]

I agree this would be awesome. Good suggestion!

…

On Fri, Jan 20, 2023 at 9:10 AM Daniel McDonald ***@***.***> wrote: It would be really cool if Qiita constructed new database-wide PCoAs on a nightly basis, and exposed the resulting .qzvs on the stats page. There are a few perspectives that may be nice: - all public - all public/private - specific studies like Microsetta For display of private data, it will be necessary to anonymize the study and sample IDs. Storing the resulting PCoAs overtime will allow for easily producing derived figures, such as the convex hull volume growth. GPU UniFrac is fast enough to be able to do this compute. And it recently was modified to enable a PCoA only output, which is valuable here as we don't need to save the resulting distance matrix. — Reply to this email directly, view it on GitHub <#3231>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ADSDCGHL2GKTO2B6QXPRUBTWTLBHNANCNFSM6AAAAAAUBX7J6Y> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>

-- Justin Shaffer, PhD Postdoctoral Researcher Rob Knight Group Department of Pediatrics, School of Medicine University of California, San Diego justinshafferbio.wordpress.com