Job Failed

[Shannon Stiles]

Hello,

I'm a new FUMA user, and I was trying to run the SNP2GENE pipeline with the MAGMA analysis in preparation for the Cell Type pipeline.  The job failed, but I did not receive an email with details on why it failed.  I was wondering if it was possible to get more information on why the job failed?  I am using summary statistics from the GWAS Catalog that I downloaded as a tsv and then zipped to compress the file to below the 600 mb limit.  I'd appreciate any help that you could provide. Thanks!

Sincerely,

Shannon Stiles

[d.p.wi...@vu.nl]

Hi Shannon, 

From your description, it sounds like the job was ended due to the time limit. If you share the jobID with me, I can look into what the specific problem was and if the job was ended due to the time limit.

Cheers,

Doug

[Mimount]

Dear Doug, 

I experience the same issue as Shannon with JobID 123232. Is it possible to also get more information on why the MAGMA analysis failed?

Thank you!

Mimount

Op dinsdag 23 maart 2021 om 09:03:12 UTC+1 schreef d.p.wi...@vu.nl:

[d.p.wi...@vu.nl]

Hi Mimount, 

I looked at job 123232 and the paramaters file indicated that option to perform MAGMA was unselected. In section six of the "New Job" page you can choose to select or unselect "Perform MAGMA". It also looks like job 123232 is OK and you should be able to access the results.

Cheers,

Doug

[Mimount]

Thanks for your fast reply! Im pretty sure I did select the MAGMA option (with 50kb gene window). But Ill try it again and see if it works now. Thank you!

Op donderdag 25 maart 2021 om 09:47:33 UTC+1 schreef d.p.wi...@vu.nl:

[Mimount]

Hi Doug, 

Im still experiencing the same issue. The email notifications tells me that the job has been completed (no errors mentioned), however the MAGMA results are not shown when I open the job (ERROR 002: MAGMA was not able to perform). 

JobID: 123456 and JobID: 123458

I made sure that I selected "Perform MAGMA". Can you see why MAGMA did not run?

Thank you again!

Mimount

Op donderdag 25 maart 2021 om 10:09:02 UTC+1 schreef Mimount:

[Shannon Stiles]

Hi Doug,

I am also still experiencing the same issues where my jobs complete with no errors, but the MAGMA analysis is not performed when it is selected.  I get the same error as Mimount (ERROR 002: MAGMA was not able to perform).  The job numbers are 123156 and 123424.  I used the default parameters for the MAGMA analysis.  I would also appreciate any information on why the MAGMA analysis is not running for this dataset.  Would you be able to provide any advice on getting the MAGMA analysis to work (in preparation for the cell type analysis) please?  Thank you!

Sincerely,

Shannon

[d.p.wi...@vu.nl]

Hi both, 

Your issues are caused by different problems.

Shannon:

Can you share the first few lines of your input file? You specify the sample size per SNP with a column label (job123156=n; job123458=N), do all your snps have a N value < 50? If so, they will be filtered out by MAGMA leading to this error "ERROR - reading p-value file: no valid p-values found for any SNPs in the data".

Mimount:

It looks like both of those jobs do not have the magma analysis selected in the parameters file. I imagine your selection of "Perform MAGMA" is not being parsed to the parameters file for some reason. Unfortunately, these problems are very hard to fix because we've been unable to recreate them so far. Often a problem parsing the selections in the input page is due to your individual browser/vpn settings. In the mean-time, do you have any queued jobs where you selected "Perform MAGMA"? If so, you can share the jobIDs with me and I will try to change the parameter file before the job starts. However, this is not a sustainable solution so I recommend trying different browsers or internet environments. 

Cheers,

Doug

[d.p.wi...@vu.nl]

Sorry Shannon, just a correction of above, both of your jobs 123156 and 123424 have "n" as the sample size column label. 

Cheers,

Doug 

[Mimount]

Hi Doug, 

Thanks a lot for checking this for me!

I do have queued some jobs now (that failed last time), so would really appreciate it if you could try to change the parameter file for me:

124299, 124300, 124301, 124305, 124306, 124321, 124323

124330, 124332, 132334, 124336, 124338, 124339, 124341

Thanks a lot for your help!

Mimount 

Op maandag 29 maart 2021 om 15:47:36 UTC+2 schreef d.p.wi...@vu.nl:

[Kristen Kelly]

Hi Mimount,

I have updated your queued jobs with the specified IDs (assuming "132334" was actually 124334), so that they will run MAGMA, using a window of 50kb (as you mentioned in an earlier reply), and using the default setting (GTEx v8) for the MAGMA tissue expression analysis.

If either the window or the tissue expression setting is not what you intended, please let me know and I can update them to something else.

Thank you,

Kristen

[Shannon Stiles]

Hi Doug and Kristen,

I replied in e-mail instead of through the FUMA group, so I'm just copying it in here now.

Thanks for your feedback on the MAGMA analysis errors.  I'm including an image of the first few lines of the input file.  All of the snps have n values greater than 50.  This is a GWAS summary statistics file for ALS from the GWAS catalog (https://www.ebi.ac.uk/gwas/studies/GCST90013429).  The original paper is by Iacoangeli et al. in 2020 (doi: https://doi.org/10.1016/j.celrep.2020.108323).  I've been trying to replicate the work in this paper to get a better understanding of using the FUMA pipeline (prior to doing this analysis I initially ran the Chron's disease example provided on the platform and had no troubles with the MAGMA analysis running). I want to use the results of the MAGMA analysis in the cell type analysis eventually.  I would appreciate any help that you could provide on how to get the MAGMA analysis working please.

Thanks for your help!

Sincerely,

Shannon

[Mimount]

Hi Kirsten, 

Thanks a lot Kirsten! All jobs have runned and now I've got the MAGMA results! 

Thanks FUMA team for your prompt help with this issue!

Mimount 

Op woensdag 31 maart 2021 om 19:59:55 UTC+2 schreef Kristen Kelly:

[Shannon Stiles]

Hi,

Over the weekend, I ran the dataset again with both the min and max gene window size. I ran it once with the MAMGA analysis gene window set to 0kb and once with the gene window set to 50 kb.  In both cases, the MAGMA analysis failed to run again.  I got the following errors for both:

Your job (job ID: 124704, job title: gwas_2020_zip_n_0kb_4-4) has been completed.

There was an error during MAGMA process (ERROR message: ERROR - reading p-value file: no valid p-values found for any SNPs in the data). This error might be because rsID in the input file did not match with MAGMA reference panel or the number of input SNPs were too small. Other results are available from the link above but MAGMA results will not be displayed.

Your job (job ID: 124705, job title: gwas_2020_zip_n_50kb_4-4) has been completed.

There was an error during MAGMA process (ERROR message: ERROR - reading p-value file: no valid p-values found for any SNPs in the data). This error might be because rsID in the input file did not match with MAGMA reference panel or the number of input SNPs were too small. Other results are available from the link above but MAGMA results will not be displayed.

This is a slightly different error than the Error 002 I received before.  In the SNP2GENE analysis, 7 genomic risk loci and 10 lead SNPs are identified.  Could you please help me figure out why I keep encountering problems with the MAGMA analysis?  Thank you!

Sincerely,

Shannon Stiles

[d.p.wi...@vu.nl]

Hi Shannon, 

Sorry for my slow reply. I donwloaded that data and it appears that the last 5 columns are just NA. I removed those 5 columns and no longer got the MAGMA error that you described. My guess as to what is happening is that the NAs get replaced by empty characters which affects the magma analysis by moving the P-value column to a column number that MAGMA is not expecting. If you remove the final 5 columns of the data then the MAGMA analysis will work fine.

Sorry for inconvenience,

Doug

[Zhiwen Jiang]

Hi Doug,

My jobs failed for MAGMA Tissue Expression Analysis. I found it depends on which datasets I selected for MAGMA gene expression analysis in section 6. In Job 206268, I selected all datasets, the MAGMA Tissue Expression Analysis failed without any output (other analysis performed as normal). In Job 206269, I used only the default datasets, GTEx v8: 30, GTEx v8: 54, it succeeded. In Job 206270, I included all the datasets except for BrainSpan29 and BrainSpan11, it succeeded. So I guess including brainspan datasets caused the failure. But my primary interest is in brain. How can I fix it? Thanks!

Best,

Owen