HMCan-Diff std::bad_alloc

[alhafidzhamdan]

Hi there, Not sure if this is the right forum but since you have another software that is related i figured i should ask here. I am trying to run HMCan-Diff using this command:

$HMCan_DIFF \
    --name ${C1}_vs_${C2}_${CHIP}
    --C1_label ${C1}_${CHIP} \
    --C2_label ${C2}_${CHIP} \
    --C1_ChIP ${C1}_${CHIP}.txt \
    --C2_ChIP ${C2}_${CHIP}.txt \
    --C1_Control ${C1}_Input..txt \
    --C2_Control ${C2}_Input.txt \
    --format BAM --pairedEnd TRUE \
    --genomePath /exports/igmm/eddie/resources/refgenome38/chromosomes \
    --GCProfile $GC_PROFILE \
    --blackListFile $ENCODE_BLACKLIST \
    --PrintWig 

$GC_PROFILE points to "GC_profile.hg38.1000bp.cnp" obtained from HMF Foundation.

I keep getting this error, despite increasing memory requirement to 100G.

HMCan-diff V1.04 is a program to identify differential regions from histone modifications in ChIP-seq cancer samples
    ..Will not use spike-in information
Consider your datasets are Paired-End...
paired end value is 1
negative binomail 0
Number of ChIP replicates for condition 1 is:2
Reading ChIP replicate 1...
/exports/igmm/eddie/ChIP-seq/SOX_TF/bams/G328_Sox2_Rep1.final.bam
Kept 68929110 reads
Reading ChIP replicate 2...
/exports/igmm/eddie/ChIP-seq/SOX_TF/bams/G328_Sox2_Rep2.final.bam
Kept 78487772 reads
Number of ChIP replicates for condition 2 is:2
Reading ChIP replicate 1...
/exports/igmm/eddie/ChIP-seq/SOX_TF/bams/G317_Sox2_Rep1.final.bam
Kept 75501471 reads
Reading ChIP replicate 2...
/exports/igmm/eddie/ChIP-seq/SOX_TF/bams/G317_Sox2_Rep2.final.bam
Kept 86926768 reads
Reading Control data for condition 1...
Kept 173442238 reads
Reading Control data for condition 2...
Kept 129248337 reads
Done reading files!!
Getting blacklist info
terminate called after throwing an instance of 'std::bad_alloc'
  what():  std::bad_alloc
/var/spool/gridscheduler/execd/node1f02/job_scripts/9088967: line 37: 27725 Aborted                 (core dumped) $HMCan_DIFF --name ${C1}_vs_${C2}_${CHIP} --C1_label ${C1}_${CHIP} --C2_label ${C2}_${CHIP} --C1_ChIP ${C1}_${CHIP}.txt --C2_ChIP ${C2}_${CHIP}.txt --C1_Control ${C1}_input.txt --C2_Control ${C2}_input.txt --format BAM --pairedEnd TRUE --genomePath /exports/igmm/eddie/resources/refgenome38/chromosomes --GCProfile $GC_PROFILE --blackListFile $ENCODE_BLACKLIST --PrintWig

How can I solve this issue? I've even subsetted my bam files to a chromosome of interest, still the same error. I've used your example files and that worked fine. I look forward to your support! Thanks in advance!

Al

[valeu]

Dear Al, Can you check that $ENCODE_BLACKLIST is not empty, file exists and the full path to the file is used? Valentina

[alhafidzhamdan]

Hi @valeu / @AshoorH

$ENCODE_BLACKLIST points to the file i downloaded directly from here - https://www.encodeproject.org/files/ENCFF356LFX/

And yes file exists and full path is used.

And genomePath points to a folder containing chr.fa and chr.fai files.

Thanks for getting back.

Al

[valeu]

Dear Al, you may use try to use the files we are using: https://github.com/BoevaLab/HMCan/tree/master/data/hg38.blacklist.bed and https://github.com/BoevaLab/HMCan/tree/master/data/GC_profile_25KbWindow_Mapp76_hg19.cnp

My apologies for a very late answer.