vg
vg copied to clipboard
Mapping paired reads w/ giraffe, no EOF marker, job stalls, exit code 79
I am mapping paired end reads to HPRC-GrCH38 (indexed with vg autoindex)
vg autoindex --workflow giraffe -x grch38.xg -g grch38.gcsa -g /path/to/file/grch38/hprc-v1.0-minigraph-grch38.gfa -p GRCH38_pangenome --target-mem 80G -t 20 > grch38_pangenome.girrafe.gbz
I am using wdl/Cromwell to execute job.; quay.io/vgteam/vg:v1.56.0; The job is stalling in the VgGiraffe alignment step
`## task order: task convertCRAMtoFASTQ task VgGiraffe task VgPack task VgCall task VgStats
task VgGiraffe { input { String sampleId File r1Fastq File r2Fastq File min File dist Int threads = 30 Int memoryGb = 200 Int diskSize File giraffe_gbz File minigraph_grch38 String outGamPath = "~{sampleId}.VgGiraffe.gam" }
command {
set -e -o pipefail
vg \
giraffe \
-t ~{threads} \
--read-group "ID:1 LB:lib1 SM:~{sampleId} PL:illumina PU:unit1" \
--sample ~{sampleId} \
--gbz-name ~{giraffe_gbz} \
--minimizer-name ~{min} \
--dist-name ~{dist} \
--fastq-in ~{r1Fastq} \
--fastq-in ~{r2Fastq} \
--ref-paths ~{minigraph_grch38} \
--progress \
--output-format gam \
--show-work \
> ~{outGamPath}
ls -thl
}`
This step produces a gam file, but the project fails with an exit code 79, so there is no EOF marker. Here are the stats from the gam file:
EOF marker is absent. The input is probably truncated Total alignments: 5118558 Total primary: 5118558 Total secondary: 0 Total aligned: 5079290 Total perfect: 3050515 Total gapless (softclips allowed): 4908914 Total paired: 5118558 Total properly paired: 5067712 Alignment score: mean 152.779, median 161, stdev 20.6774, max 161 (3050515 reads) Mapping quality: mean 54.2928, median 60, stdev 16.4384, max 60 (4427530 reads) Insertions: 272418 bp in 93519 read events Deletions: 303806 bp in 106883 read events Substitutions: 5294946 bp in 5294946 read events Softclips: 12320666 bp in 264475 read events Total time: 1.75599e+06 seconds Speed: 2.91492 reads/second
The docker was successfully pulled, and with these stats on the gam file, I don't understand why the job fails to go to the next task. I did a dry run on the wdl itself and that was successful, so no syntax errors.