Segmentation failed (core dumped) error in featureCounts

Question

[Batu]

Hello,

When I run featureCounts, it gives an error after loading GTF file and showing number of chromosomes.

    featureCounts -T 3 -t exon -g gene_id -a "/mnt/c/Users/user/Desktop/STEMCELL/RNA_3/References/Homo_sapiens.GRCh38.92.gtf" -o "/mnt/c/Users/user/Desktop/STEMCELL/RNA_3/Output_Files/SRR1639080_readCounts.txt" "/mnt/c/Users/user/Desktop/STEMCELL/RNA_3/Output_Files/SRR1639080_sorted.bam"

Output is below:

        ==========     _____ _    _ ____  _____  ______          _____
        =====         / ____| |  | |  _ \|  __ \|  ____|   /\   |  __ \
          =====      | (___ | |  | | |_) | |__) | |__     /  \  | |  | |
            ====      \___ \| |  | |  _ <|  _  /|  __|   / /\ \ | |  | |
              ====    ____) | |__| | |_) | | \ \| |____ / ____ \| |__| |
        ==========   |_____/ \____/|____/|_|  \_\______/_/    \_\_____/
          v1.6.0

//========================== featureCounts setting ===========================\\
||                                                                            ||
||             Input files : 1 BAM file                                       ||
||                           S /mnt/c/Users/user/Desktop/STEMCELL/RNA_3/O ... ||
||                                                                            ||
||             Output file : /mnt/c/Users/user/Desktop/STEMCELL/RNA_3/Out ... ||
||                 Summary : /mnt/c/Users/user/Desktop/STEMCELL/RNA_3/Out ... ||
||              Annotation : /mnt/c/Users/user/Desktop/STEMCELL/RNA_3/Ref ... ||
||      Dir for temp files : /mnt/c/Users/user/Desktop/STEMCELL/RNA_3/Out ... ||
||                                                                            ||
||                 Threads : 3                                                ||
||                   Level : meta-feature level                               ||
||              Paired-end : no                                               ||
||         Strand specific : no                                               ||
||      Multimapping reads : not counted                                      ||
|| Multi-overlapping reads : not counted                                      ||
||   Min overlapping bases : 1                                                ||
||                                                                            ||
\\===================== http://subread.sourceforge.net/ ======================//

//================================= Running ==================================\\
||                                                                            ||
|| Load annotation file /mnt/c/Users/user/Desktop/STEMCELL/RNA_3/Referenc ... ||
||    Features : 1237303                                                      ||
||    Meta-features : 58395                                                   ||
||    Chromosomes/contigs : 47                                                ||
||                                                                            ||
Segmentation fault (core dumped)

I couldn't find anything specific for featureCounts. But in the posts, it is generally said that "Segmentation fault" means that you tried to access memory that you do not have access to. I couldn't resolve how I can resolve it.

Any ideas or suggestions? Thanks...

[Devon Ryan]

I've occasionally run into segfaults with featureCounts and found that switching versions (typically to the newest one) tends to resolve this.

[Batu]

Switching with 1.6.2 did not resolve the problem

[ioannis]

If anyone is using multiple BAM files with featureCounts and gets the exact same error, it's because of the multicore function (-T). Remove -T option and it will work like a charm with multiple BAM files.

Answer 1

[Batu]

Workaround for this problem is to recompile subread package with the recent glibc (https://github.com/Microsoft/WSL/issues/3391). To do that, update glibc to the newest version (2.23), and then compile the source of the package, and introduce its "bin" directory to path. Problem solved!!

Answer 2

[ATpoint]

How much memory do you have available?

[Batu]

I have 8GB RAM on my PC

[Batu]

I've tried the same code in another PC with 16GB of memory, and it worked seamlessly. So the problem was the memory. Thanks for your help!!

[ATpoint]

Good to hear it worked. Please consider to mark the answer as accepted to help others in the future.

[Batu]

My friend tried featureCounts on a computer with much higher RAM, but she got the same error. When I solved the problem, I tried on computer with Ubuntu. But me and my friend are using Ubuntu in Windows (WSL). Can this be a reason? How and why this problem occurs? Any idea? Thanks...

[Devon Ryan]

Hard to say, upload the files to a public Galaxy server and try there. That'll be using a regular linux server to run things.

[ATpoint]

It does not matter how much RAM the host computer has, but how much is allocated to the virtual machine. Check how much is indeed consumable for the VM.

[Batu]

Actually, WSL is not a virtual machine. It is a subsystem in Windows and works almost natively (almost, because I had an error). I saw that there was a bug that restricts the usage of RAM on WSL 2 years ago, but now it is fixed and I can see all amount of RAM there.