HippocampusGirl / ImputationProtocol

The ENIGMA Imputation Protocol as a container for your local workstation or high-performance compute cluster
2 stars 4 forks source link

Error installing imputation server #3

Open HippocampusGirl opened 7 months ago

HippocampusGirl commented 7 months ago

image

nadineparker commented 5 months ago

I get a similar error when running step #5 setup-imputationserver. We work on a closed system that does not have external internet access. I assume this is an issue for downloading the 1000G reference data? Is there a possible work around for this?

Screenshot 2024-01-28 at 20 54 27
HippocampusGirl commented 5 months ago

Hi @nadineparker,

Could you confirm when you downloaded the container? I uploaded a new version last Thursday (January 25) that I tested without internet access locally, and it seems to work for me.

If you downloaded before then, the issue might simply be fixed by re-downloading. Otherwise, I will have another look and test again :-)

HippocampusGirl commented 5 months ago

Oh no, my bad, I see the problem.

I only tested with Docker, but Singularity seems to be failing due to a permission error. This should be fixed now!

nadineparker commented 5 months ago

Thanks! Using the updated container I was able to run the "setup-imputationserver". However, now I have an issue at step 8 when using the imputationserver command. Loading a particular R library (geneplotter) seems to fail. Note that I'm using the eur population. Here is a screen shot of the error. Screenshot 2024-01-29 at 23 58 11

HippocampusGirl commented 5 months ago

Hi @nadineparker, I have only been testing with the mixed population, which doesn't need R at all, so I never saw this problem! I will try to upload a fix as soon as possible

nadineparker commented 5 months ago

Hi, I ended up using the mixed population as well and I eventually got everything to run. I'm curious about the criteria for sample missingness and subsequent chunk exclusions? It seemed as though having one person who surpassed the missingness threshold resulted in the entire chunk being excluded is this supposed to be the case? and what is the missingness threshold? Sorry for all the questions just wanted to understand the pipeline a bit better. Thanks for all of your help!