BCL2FASTQ: Segmentation Fault Error (segfault)

Question

[gillhuberfelix]

Hi all,

I need to preprocess my data with bcl2fastq. Unfortunately, no matter how I try it, I get a "Segmentation Fault" error. I call bcl2fastq with the command bcl2fastq --no-lane-splitting --mask-short-adapter-reads=13. My data structure is shown below. I thought I had to call the program on the ~/Data/Raw Data/170203_NB501802_0007_AH5YHJBGX2$ level, but maybe I'm mistaken. However, I'll get the same error if I try to call bcl2fastq on another level. If I try to define the input directory with the --input-dir, I'll get the error: Failed to parse the options: too many positional options have been specified on the command line. I appreciate any advice on how to solve this issue!

[GenoMax]

What OS are you trying to do this on? Please provide full command line (you can remove specific paths if you want to) that you are using.

You minimally need to provide

bcl2fastq -R /path_to_flowcell -o /path_to_dir_to_write_results --sample-sheet SampleSheet.csv -r N -p N -w N

N = number of cores to use for read/processing/write operations.

[gillhuberfelix]

Hey,

Thanks for your reply.

I use Linux Debian 6.1.55-1.

I filled out the command you provided:

bcl2fastq -R ./170203_NB501802_0007_AH5YHJBGX2/Data/Intensities/BaseCalls/ -o ~/home/felix/output/ --sample-sheet ./170203_Samples_MUC-05_VS02.csv -r 2 -p 12 -w 2

But I get the same error, or have I got something wrong?

[GenoMax]

How did you install the program? Did you compile it from source?

Please try the following command. You have to point to the top level of the flowcell folder for -R.

bcl2fastq -R ./170203_NB501802_0007_AH5YHJBGX2/ -o ~/home/felix/output/ --sample-sheet ./170203_Samples_MUC-05_VS02.csv -r 2 -p 4 -w 2

If you are really using debian 6 then that is an ancient OS from 2011.

[gillhuberfelix]

I installed the program with the .rpm package and used alien. I tried your command, but unfortunately, it doesn't work either.

Sorry, I think I got something wrong with the OS version. The server runs on Devuan version Daedalus 5.

[GenoMax]

Are you able to see inline help if you just try to run bcl2fastq -h? If you get the same error they you will need to compile the program from source to get it to work on your OS.

[gillhuberfelix]

Yeah, I am able to see the help section, and I can also run bcl2fastq -v. So, I think the installation process was successful.

[GenoMax]

If you are still getting "seg faults" perhaps other possibility is that you don't have (or are not assigning) enough memory.

[gillhuberfelix]

I think I have enough memory:

[GenoMax]

It looks like there is some interaction between alien/your OS/bcl2fastq code that is causing this issue. If you don't have another option then you may want to try compiling from source.

I notice that this is an old flowcell. Perhaps you could try to use GATK BasecallsToSam as alternative: https://gatk.broadinstitute.org/hc/en-us/articles/360037225472-IlluminaBasecallsToSam-Picard-

[Darked89]

Which version of bcl2fastq are you using?

If I remember correctly the newest bcl files are not compatible with older versions of bcl2fastq.

Also: try to download some minimal dataset for bcl2fastq2 and check if your bcl2fastq is able to process it.

[GenoMax]

Looking at the flowcell folder ID this run is likely from 2017.

Answer 1

[gillhuberfelix]

Thank you both for your help.

I fixed it. The problem was with the Linux distribution. I installed it again with Conda, and that worked.