Purple doesnt run without structural variant input

[SebastianHollizeck]

Hi,

i was testing purple on our data and wanted to try a few thinsg while waiting for gridss results to be completed.

when running purple with just

java -jar /dawson_genomics/Other/software/purple/purple_v3.1.jar -reference CA80 -tumor CA80-2 -amber ./amber -cobalt ./cobalt -ref_genome /data/reference/dawson_labs/genomes/GRCh38/GCA_000001405.15_GRCh38_full_analysis_set.fna -gc_profile /dawson_genomics/Other/software/cobalt/GC_profile.1000bp.38.cnp -circos circos -output_dir .

i get the message

11:11:48 - [INFO ] - PURPLE version: 3.1
11:11:48 - [INFO ] - Reference Sample: CA80, Tumor Sample: CA80-2
11:11:48 - [INFO ] - Output Directory: ./
11:11:48 - [INFO ] - Using ref genome: V38
11:11:49 - [INFO ] - Reading GC Profiles from /dawson_genomics/Other/software/cobalt/GC_profile.1000bp.38.cnp
11:11:51 - [INFO ] - Processing sample(ref=CA80 tumor=CA80-2)
11:11:51 - [ERROR] - failed processing sample(CA80-2): org.apache.commons.cli.ParseException: missing structural_vcf or sample data directory
org.apache.commons.cli.ParseException: missing structural_vcf or sample data directory
    at com.hartwig.hmftools.purple.config.SampleDataFiles.getFilename(SampleDataFiles.java:79)
    at com.hartwig.hmftools.purple.config.SampleDataFiles.<init>(SampleDataFiles.java:63)
    at com.hartwig.hmftools.purple.PurpleApplication.processSample(PurpleApplication.java:196)
    at com.hartwig.hmftools.purple.PurpleApplication.run(PurpleApplication.java:150)
    at com.hartwig.hmftools.purple.PurpleApplication.main(PurpleApplication.java:503)
11:11:51 - [INFO ] - Complete

both of those inputs are flagged as optional on the help page.

Both 3.0 and 3.1 have this behaviour, but 2.54 seems to work without svs

Cheers, Sebastian

[charlesshale]

I've restored this functionality. In Purple 3.2 beta:

https://github.com/hartwigmedical/hmftools/releases/tag/purple-v3.2_beta

[teng-gao]

Hi,

Thanks for this amazing pipeline!

I ran into the below error when running Purple-v3.2-beta. My command:

java -jar ~/hartwig/purple.jar \
        -tumor_only \
        -tumor $sample \
        -output_dir ~/external/WASHU/$sample/purple \
        -amber ~/external/WASHU/$sample/amber \
        -cobalt ~/external/WASHU/$sample/cobalt \
        -gc_profile ~/hartwig/GC_profile.1000bp.38.cnp \
        -ref_genome ~/ref/hg38.fa

The error message:

22:52:50 - [INFO ] - PURPLE version: 3.2
22:52:50 - [INFO ] - Tumor Sample: 47491_Primary
22:52:50 - [INFO ] - Output Directory: /home/tenggao/external/WASHU/47491_Primary/purple/
22:52:50 - [INFO ] - Using ref genome: V37
22:52:51 - [INFO ] - Reading GC Profiles from /home/tenggao/hartwig/GC_profile.1000bp.38.cnp
22:52:53 - [INFO ] - Processing sample(ref=DIPLOID tumor=47491_Primary)
22:52:53 - [INFO ] - Reading amber QC from /home/tenggao/external/WASHU/47491_Primary/amber/47491_Primary.amber.qc
22:52:53 - [INFO ] - Reading amber bafs from /home/tenggao/external/WASHU/47491_Primary/amber/47491_Primary.amber.baf.tsv
22:52:54 - [INFO ] - Reading amber pcfs from /home/tenggao/external/WASHU/47491_Primary/amber/47491_Primary.amber.baf.pcf
22:52:54 - [INFO ] - Average amber tumor depth is 53 reads implying an ambiguous BAF of 0.561
22:52:54 - [INFO ] - Reading cobalt ratios from /home/tenggao/external/WASHU/47491_Primary/cobalt/47491_Primary.cobalt.ratio.tsv
22:52:57 - [INFO ] - Reading cobalt reference segments from /home/tenggao/external/WASHU/47491_Primary/cobalt/DIPLOID.cobalt.ratio.pcf
22:52:57 - [INFO ] - Reading cobalt tumor segments from /home/tenggao/external/WASHU/47491_Primary/cobalt/47491_Primary.cobalt.ratio.pcf
22:52:57 - [INFO ] - Somatic variants support disabled.
22:52:57 - [INFO ] - Sample gender is male
22:52:57 - [INFO ] - Applying segmentation
22:52:57 - [INFO ] - Merging reference and tumor ratio break points
22:53:00 - [INFO ] - Fitting purity
22:53:22 - [INFO ] - Sample maxDiploidProportion(0.567) diploidCandidates(93) purityRange(0.800 - 0.890) hasTumor(true)
22:53:22 - [INFO ] - Calculating copy number
22:53:23 - [INFO ] - Generating QC Stats
22:53:23 - [INFO ] - Modelling somatic peaks
22:53:23 - [ERROR] - failed processing sample(47491_Primary): htsjdk.tribble.TribbleException$MalformedFeatureFile: Unable to parse header with error: java.io.IOException: Is a directory, for input source: file:///home/tenggao/
htsjdk.tribble.TribbleException$MalformedFeatureFile: Unable to parse header with error: java.io.IOException: Is a directory, for input source: file:///home/tenggao/
        at htsjdk.tribble.TribbleIndexedFeatureReader.readHeader(TribbleIndexedFeatureReader.java:263)
        at htsjdk.tribble.TribbleIndexedFeatureReader.<init>(TribbleIndexedFeatureReader.java:102)
        at htsjdk.tribble.TribbleIndexedFeatureReader.<init>(TribbleIndexedFeatureReader.java:127)
        at htsjdk.tribble.AbstractFeatureReader.getFeatureReader(AbstractFeatureReader.java:121)
        at htsjdk.tribble.AbstractFeatureReader.getFeatureReader(AbstractFeatureReader.java:81)
        at htsjdk.variant.vcf.VCFFileReader.<init>(VCFFileReader.java:145)
        at htsjdk.variant.vcf.VCFFileReader.<init>(VCFFileReader.java:95)
        at com.hartwig.hmftools.purple.somatic.SomaticPeakStream.somaticPeakModel(SomaticPeakStream.java:61)
        at com.hartwig.hmftools.purple.PurpleApplication.performFit(PurpleApplication.java:315)
        at com.hartwig.hmftools.purple.PurpleApplication.processSample(PurpleApplication.java:223)
        at com.hartwig.hmftools.purple.PurpleApplication.run(PurpleApplication.java:156)
        at com.hartwig.hmftools.purple.PurpleApplication.main(PurpleApplication.java:563)
Caused by: htsjdk.samtools.util.RuntimeIOException: java.io.IOException: Is a directory
        at htsjdk.tribble.readers.SynchronousLineReader.readLine(SynchronousLineReader.java:53)
        at htsjdk.tribble.readers.LineIteratorImpl.advance(LineIteratorImpl.java:24)
        at htsjdk.tribble.readers.LineIteratorImpl.advance(LineIteratorImpl.java:11)

Note that besides this error, the gnome version is also not recognized (I passed hg38).

[charlesshale]

I've restored Purple's ability to run without a somatic VCF. The crash above is due to Purple assuming the file has been provided in config.

Also you'll need to specify the config "-ref_genome_version V38" - this is now in the example config in the read-me:

Beta JAR updated:

https://github.com/hartwigmedical/hmftools/releases/tag/purple-v3.2_beta