Assistance Needed with FUMA Cell Type Analysis Error (Job ID: 682360, 682361, 682362)

[Jie Sun]

Dear FUMA Development Team,

I am writing to seek assistance with an error I encountered during the cell type specificity analysis in FUMA. Below are the details of my jobs:

Job ID: 682360, 682361, 682362

Job Title: CELL, BRAIN, METABOLISM

The analysis failed with the following error message:

"ERROR: cellType - An error occurred during the process of your cell type specificity analyses."

Could you please help me identify the issue? If needed, I can provide the input files for further inspection.

Thank you for your time and support. I look forward to your response.

Best regards,

Jie Sun

[Tanya Phung]

Hi Jie, 

The errors of your jobs are similar to the ones answered here: https://groups.google.com/g/fuma-gwas-users/c/W7WFq834TVc/m/aY9IQwP_BQAJ

However, these are the problematic datasets:

job 682360

• 539_Travaglini_2020_Blood_level1

job 682361

• GSE89232_Human_Blood

job 682362

• 272_Aldinger2021_PrenatalCerebellum_Human_2021_21wpc_level1

Maybe you can try again with excluding this dataset from your analysis. 

Best,
Tanya

[Jie Sun]

Dear Tanya,

Following your suggestion, I excluded the three problematic datasets (539_Travaglini_2020_Blood_level1, GSE89232_Human_Blood, and 272_Aldinger2021_PrenatalCerebellum_Human_2021_21wpc_level1) and reran the cell type analyses (Job IDs: 683583, 683582, 683581).

However, the same "ERROR: cellType" persists.

Could you please advise on the next steps?

Is there a known issue with the cell type analysis module for specific gene sets or datasets?

Would sharing the input file or job logs help diagnose the problem?

Should I try rerunning the SNP2GENE step or switch to a different gene ID format (e.g., Entrez IDs)?

Thank you for your support. I look forward to your guidance.

Jie Sun

[Tanya Phung]

Hi Jie, 

In FUMA Cell type module, MAGMA is implemented. MAGMA is designed to run on genome-wide data. You used the magma results from your SNP2GENE job 682252 where there are only 22 genes. So the error of the jobs are: no valid gene sets or covariates found in input files. This is because MAGMA is designed to run with files with typically >10,000 genes. 

Best,
Tanya

[Jie Sun]

Dear Tanya,

Thank you again for your previous explanation regarding the MAGMA error in my SNP2GENE job (ID: 682252). I now understand that the cell-type analysis using MAGMA requires a genome-wide background and is not suitable for my small set of 22 candidate genes.

My primary goal is to perform cell-type enrichment analysis specifically for these 22 genes within the FUMA platform.

Could you kindly guide me on the correct operational steps? Specifically:

1. Is it possible to use the GENE2FUNC module with my small gene list to obtain cell-type specificity results, even if statistical enrichment tests (like MAGMA) are not applicable?

2. If the above is not feasible, could you please advise on the recommended workflow or alternative functions within FUMA to explore the cell-type expression patterns or relevance of my prioritized gene list?

Thank you for your time and support. I greatly appreciate your help.

Best regards,
Jie