Pipeline failure on second read during bowtie2_end_to_end

[Rensco Hogers]

Hey all,
I encountered the following problem:

Check Documentation

I have checked the following places for this error:

• nf-core website: troubleshooting
• nf-core/hic pipeline documentation

Description of the bug

When running the pipeline, everything seems to be fine and the first few tasks are completed, from getting the sofware version to getting the restriction fragments. It goes wrong when an end to end allignment is created used bowtie2. It finishes fairly quickly for the first read file and then proceeds to the trimming of the reads, performing another bowtie2 run on it and then merging of mapping steps. But then it never finishes with the second read file for the end to end alignment. I have tried using my filtered dataset, for which fastp was used to filter out low quality reads, as well as the raw Hi-C reads.

Below I have posted the progress it will make before failing out due to the second alignment taking too long, it will stop after eight hours. I have also attached the logfile it produces.

Is there something that I have done wrong, or any tips you can give me to complete the pipeline successfully?

[32/d2db3b] process > get_software_versions [100%] 1 of 1 ✔
[a2/fd33ca] process > makeBowtie2Index (long-tailed_finch_hi-c.p_ctg) [100%] 1 of 1✔
[1c/2be66e] process > makeChromSize (long-tailed_finch_hi-c.p_ctg.fa) [100%] 1 of 1 ✔
[ff/ca8c74] process > getRestrictionFragments (long-tailed_finch_hi-c.p_ctg.fa ^GATC,G^ANTC) [100%] 1 of 1✔
[fd/bb2213] process > bowtie2_end_to_end (long-tailed-finch_R1_filtered_R1) [ 50%] 1 of 2
[40/3f9945] process > trim_reads (long-tailed-finch_R1_filtered_R2) [100%] 1 of 1
[ff/2bc982] process > bowtie2_on_trimmed_reads (long-tailed-finch_R1_filtered_R2) [100%] 1 of 1
[ae/047d1e] process > bowtie2_merge_mapping_steps (long-tailed-finch_R1_filtered_R2 = input.1.bam + input.1_unmap_trimmed.bam) [100%] 1 of 1
[- ] process > combine_mates -
[- ] process > get_valid_interaction -
[- ] process > remove_duplicates -
[- ] process > merge_stats -
[- ] process > build_contact_maps -
[- ] process > run_ice -
[- ] process > convert_to_pairs -
[- ] process > cooler_raw -
[- ] process > cooler_balance -
[- ] process > cooler_zoomify -
[- ] process > dist_decay -
[- ] process > compartment_calling -
[- ] process > tads_hicexplorer -
[- ] process > tads_insulation -
[- ] process > multiqc -
[57/58fac6] process > output_documentation [100%] 1 of 1 ✔

Steps to reproduce

Steps to reproduce the behaviour:

1. Command line:
   ./nextflow run nf-core/hic -profile docker --input ../../Data/hi-c_reads/processed_reads/*.fastq.gz --outdir ../../Data/hi-c_reads/analysis
   --fasta ../../Data/hifiasm_results/long-tailed_finch_hi-c/long-tailed_finch_hi-c.p_ctg.fa --restriction_site '^GATC,G^ANTC' --ligation_site 'GATCGATC,GATCANTC,GANTGATC,GANTANTC' --max_cpus 64 --max_memory 600GB --bwt2_opts_end2end '--very-sensitive -L 30 --score-min L,-0.6,-0.6 --end-to-end --reorder -p 64'

Expected behaviour

I expected all the tasks to be completed, so first the one read file would be alligned using bowtie2, and then the second one would.

Log files

Have you provided the following extra information/files:

• The command used to run the pipeline
• The .nextflow.log file

System

• Hardware: HPC
• Executor: local
• OS: Ubuntu
• Version: 18.04.6

Nextflow Installation

• Version: 22.04.3

Container engine

• Engine: Docker
• version: 20.10.16
• Image tag: nfcore/hic:1.3.0

nextflow log.log