Impute variants from fastq files - homozygous error

Question

[Bioin]

Hi everyone, I was trying imputing variants from .fq files using the below given command:

docker run --name pipeline_Impute --rm \
    -v /data/Projects/WGRS/PHG/Imputation/:/tempFileDir/data/fastq/ \
    -v /data/Projects/WGRS/PHG/1_buildPHG/:/tempFileDir/data/ \
    -v /data/Projects/WGRS/PHG/DockerOutput/PHG_292taxa.db:/tempFileDir/outputDir/PHG_292taxa.db \
    -v /data/Projects/WGRS/PHG/DockerOutput/PangenomeFasta/:/tempFileDir/outputDir/pangenome/ \
    -t maizegenetics/phg:latest /tassel-5-standalone/run_pipeline.pl -Xmx50G -debug -configParameters /tempFileDir/data/Config_Impute_fq.txt -ImputePipelinePlugin -imputeTarget path -endPlugin

Unfortunately, I was prompted by an error repeatedly. Please find below the last lines of the log together with the error:

........above lines were trimmed.............

DefaultDispatcher-worker-5] DEBUG net.maizegenetics.pangenome.api.FilterGraphPlugin - Filter graph on refRanges.
[DefaultDispatcher-worker-5] INFO net.maizegenetics.pangenome.api.CreateGraphUtils - createEdges: creating edges from nodes.
[DefaultDispatcher-worker-5] INFO net.maizegenetics.pangenome.api.CreateGraphUtils - createEdges: time: 6.017400259 secs.
[DefaultDispatcher-worker-5] INFO net.maizegenetics.pangenome.api.HaplotypeGraph - Created graph number of edges: 1066384  number of nodes: 1068720  number of reference ranges: 3660
[DefaultDispatcher-worker-5] DEBUG net.maizegenetics.pangenome.hapCalling.ConvertReadsToPathUsingHMM - after filtering hapgraph: 1068720 nodes.

**[DefaultDispatcher-worker-5] INFO net.maizegenetics.pangenome.hapCalling.ConvertReadsToPathUsingHMM - Getting path for chromosome 7
[pool-1-thread-1] DEBUG net.maizegenetics.plugindef.AbstractPlugin - -16**
**java.lang.ArrayIndexOutOfBoundsException: -16**
    at net.maizegenetics.analysis.imputation.ViterbiAlgorithmVariableStateNumber.getMostProbableStateSequence(ViterbiAlgorithmVariableStateNumber.java:180)
    at net.maizegenetics.pangenome.hapCalling.ConvertReadsToPathUsingHMM.haplotypeCountsToPath(ConvertReadsToPathUsingHMM.java:281)
    at net.maizegenetics.pangenome.hapCalling.BestHaplotypePathPlugin$processKeyFileEntry$2.invokeSuspend(BestHaplotypePathPlugin.kt:539)
    at kotlin.coroutines.jvm.internal.BaseContinuationImpl.resumeWith(ContinuationImpl.kt:33)
    at kotlinx.coroutines.DispatchedTask.run(Dispatched.kt:241)
    at kotlinx.coroutines.scheduling.CoroutineScheduler.runSafely(CoroutineScheduler.kt:594)
    at kotlinx.coroutines.scheduling.CoroutineScheduler.access$runSafely(CoroutineScheduler.kt:60)
    at kotlinx.coroutines.scheduling.CoroutineScheduler$Worker.run(CoroutineScheduler.kt:740)
[pool-1-thread-1] INFO net.maizegenetics.plugindef.AbstractPlugin - 
BestHaplotypePathPlugin Description...
Plugin that takes a haplotype graph and a set of read mappings to infer the best path through the graph. The input read mappings can be supplied as a file, a directory of files, or entries from a DB.

Usage:
BestHaplotypePathPlugin <options>
.........options
.....
..etc

**[pool-1-thread-1] ERROR net.maizegenetics.plugindef.AbstractPlugin - -16**

what should I do to solve this problem, could anyone give me some advice ？Thanks in advance.

Answer 1

[pjb39]

One possible cause is that the number of haplotypes for one or more reference ranges is greater than 127. If that is the case, then using consensus haplotypes should reduce the number below that. To make sure it does, you can add the parameter maxClusters to the config file used to create consensus haplotypes and set that to a number less than 128. An alternative would be to limit the number of taxa used for imputation.

[Bioin]

As suggested, I ran the create consensus haplotypes with maxClusters set to 127 even then I end up with the same error message. Any help on this would be appreciated. Thank you.