Job 359393 (minor-ish) error

Viktor

unread,

Jul 2, 2021, 1:51:50 PM7/2/21

to GenePattern Help Forum

Hello,

I am getting the following error message from ComparativeMarkerSelection job: "Unable to compute q values."

Apart from q-values which are all NaN, the job seems to have completed successfully, All other statistics, including FDR (BH), are computed and present in the output table.

Please advise and thank you,

Viktor

unread,

Jul 6, 2021, 10:17:05 AM7/6/21

to GenePattern Help Forum

Still waiting for help here.

Barbara Hill

unread,

Jul 6, 2021, 11:35:46 AM7/6/21

to GenePattern Help Forum

Hello Viktor,

Apologies for the wait - the team was all out of office for the national holiday.

It looks like this module was never configured to run in a container that has R 2.5 which is what is causing this error. IE the java portion of the code is running, but R 2.5 is needed to comput q values and there is no R in the default Java 7 container.

We'll keep you posted as we work to fix this.

Best

-Barbara

Viktor

unread,

Jul 6, 2021, 11:37:27 AM7/6/21

to GenePattern Help Forum

Hi Barbara,

No worries and thank you for looking into this.

Best,

Viktor

Barbara Hill

unread,

Jul 27, 2021, 5:08:37 PM7/27/21

to GenePattern Help Forum

Hi Viktor,

I've updated ComparativeMarkerSelection and it is once again computing q values.

Please let us know if you run into anything else.

Best

-Barbara

Viktor

unread,

Jul 27, 2021, 6:19:09 PM7/27/21

to GenePattern Help Forum

Hi Barbara,

Perfect, thank you very much.

Best,

Viktor

Reply all

Reply to author

Forward