Is the extended distance bin (nearly 1 Mb from gene body) is supported by sQTLseekr2 for calling the sQTLs ?

[gauriprasad08]

Dear Developers,

I wanted to reach out regarding an issue I encountered while running the sqtlseekR2 analysis on the Biowulf server at NIH.

While using a distance bin of +1 Mb from the gene body through the nextflow setup, I encountered the following error message:

"*** caught segfault *** address 0x18, cause 'memory not mapped'."

I have attached the full error and my batch script for your reference.

Interestingly, when I reduced the distance bin to only 20kb, the analysis ran smoothly and produced the expected outputs from the same input files. However, for final analysis, I am keen to utilize the 1Mb distance bin.

Could you kindly advise on any parameters that I might need to adjust to enable the analysis to run successfully with the 1Mb distance bin? Additionally, I would appreciate your insights on whether the pipeline is recommended or well suited for running with a 1Mb distance bin.

Thank you very much for your guidance on this matter. Your expertise and support are greatly appreciated.

Best,
Gauri
sqtlseekR2_nextflow_run_1Mb_script.txt
Log_file.txt

[dgarrimar]

Please note that 1 Mb is a pretty large window (in sQTLseekeR the window is not defined flanking the TSS, but rather flanking the whole gene). In addition, for splicing analyses, a much smaller window would be more than enough. Do you still get the error when using a smaller window size, e.g. 5Kb?

[gauriprasad08]

Hi

Thank you very much for the response.

I can successfully get my outputs without any errors when I use a distance window of 5 Kb.

Best,
Gauri

[dgarrimar]

it seems just a memory problem, probably increasing the memory that each job is allocated it will work