writeVcf save vcf with invalid contig information

[thomasfaraut]

If I read a valid vcf with the last 3.10 Bioconductor version 1.32.0

• when the vcf file contains a single contig, the ouput contig information is truncated and not valid

vcf <- readVcf("input_singlecontig.vcf")
writeVcf(vcf, "output_singlecontig.vcf")

head of input_singlecontig.vcf

##fileformat=VCFv4.2
##contig=<ID=10,length=135534747>
##ALT=<ID=INV,Description="Inversion">

head of output_singlecontig.vcf

##fileformat=VCFv4.2
##fileDate=20200109
##FILTER=<ID=PASS,Description="All filters passed">
##FILTER=<ID=LOW_QUAL,Description="Low quality call as specified by 'variantcalling.lowQuality'">
##contig=135534747

• but works correctly when at least two contigs are present, e.g.

  ##contig=<ID=10,length=135534747>
  ##contig=<ID=11,length=135006516>

  I am attaching the two files so that the error can be reproduced easily (extensions changed to .txt): input_singlecontig.txt output_singlecontig.txt

Thanks for you help!

[lshep]

Hello. Thank you for the report. I'm following up on this now. Can you provide the input file with two contigs so I can compare the output from valid to invalid?

[lshep]

Is this still an issue?

[lshep]

With no follow up information we are closing this issue. If this i still an issue please request it be reopened with the requested additional information

[thomasfaraut]

Sorry for my late reaction. I missed the first two emails because I dit not recognize it was related to the submitted issue.

The input file with two contigs could have simply be constructed by adding the line

 ##contig=<ID=11,length=135006516>

to the provided input_singlecontig.txt.

I didn't answered earlier to your two last emails because I encountered troubles to upgrade my Bioconductor installation (I tried even to reinstall) to be able to see if it was still an issue. I am not using anymaore Bioconductor on a regular basis, this explains why it was difficult to me to reproduce the error.

I apologize again.

[lshep]

No worries. Thank you for the additional information and letting us know it is still an issue.

[Lenbok]

Seems to me that this issue and #42 probably have exactly the same cause.