Differentiate between errors and restricted datasets

[bschilder]

Some OpenGWAS datasets don't show any data links on their website, I think because they're restricted use (at least currently). e.g. https://gwas.mrcieu.ac.uk/datasets/finn-b-G6_ALZHEIMER/

Currently, MungeSumstats gives a message saying this this might be due to an incorrect ID or URL. However, it would be nice to let the user know it's simply restricted and won't be downloaded.

We could either do this:

• in the beginning of the import_sumstats step, or
• the end of find_sumstats step, and provide a message to users that X, Y, Z GWAS were dropped from the results due to lack of accessibility (default). If so, we can add an arg to just show all results regardless of accessibility (non-default).

On that note, I think there is a way you can provide your OpenGWAS access token and get access to additional GWAS you've been authorized for. We just need to pass up the arg to supply the token to import_sumstats.

[bschilder]

Just checked if the OpenGWAS metadata can be used for this. They give a column called "group_name", which at first glance seems to be what I want.

But then i realised they always seem to be set to "public", even for the example above that we know is missing.

[bschilder]

Regarding the access token, i just remembered that we don't use OpenGWAS's code for downloading VCFs, bc ieugwasr/gwasvcf only let you query a small subset of a given VCF, not download the entire VCF (at least last time I checked).
That's why i made all the downloader functions and constructed the VCF URLs inside MungeSumstats.

So I'm not sure if there's currently an API-accessible means of downloading private VCF from OpenGWAS. Something to discuss with their team.

[NathanSkene]

If they have a system for managing permissions, would make sense to incorporate it

…

________________________________ From: Brian M. Schilder ***@***.***> Sent: 17 November 2021 21:03 To: neurogenomics/MungeSumstats ***@***.***> Cc: Subscribed ***@***.***> Subject: Re: [neurogenomics/MungeSumstats] Differentiate between errors and restricted datasets (Issue #67) This email from ***@***.*** originates from outside Imperial. Do not click on links and attachments unless you recognise the sender. If you trust the sender, add them to your safe senders list<https://spam.ic.ac.uk/SpamConsole/Senders.aspx> to disable email stamping for this address. Regarding the access token, i just remembered that we don't use OpenGWAS's code for downloading VCFs, bc ieugwasr/gwasvcf only let you query subset of the VCFs, not download the entire VCF (at least last time I checked). That's why i made all the downloader functions and constructed the VCF URLs inside MungeSumstats. So I'm not sure if there's currently an API-accessible means of downloading private VCF from OpenGWAS. Something to discuss with their team. — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub<https://github.com/neurogenomics/MungeSumstats/issues/67#issuecomment-972020451>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/AH5ZPEYXU7ESBDOPL7NHTTTUMQKA3ANCNFSM5IFXOCXA>. Triage notifications on the go with GitHub Mobile for iOS<https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android<https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>.

[Al-Murphy]

My understanding was we couldn't use it since the package isn't on CRAN/Bioconductor. @bschilder can you confirm?

[bschilder]

Yeah, there were two reasons:

1. Exactly as @Al-Murphy mentioned, these are not CRAN/Bioc-distributed, so we can't make them deps for MungeSumstats (a Bioc restriction). The way i got the metadata search functions to work was by copying and pasting the relevant code.
2. Neither of the packages allows you to download whole-VCF files, which is what we need for MungeSumstats