[Bioc-devel] readVcf problems in package check

[Clark, Lindsay V]

Hi Herve,

Thanks so much for your response!  I realize that you all aren't obligated to support CRAN packages.  When I was developing polyRAD I considered submitting Bioconductor, but I didn't want to have to convert my 'RADdata' class from S3 to S4.  The polyRAD package also isn't very deeply integrated with Bioconductor and doesn't require users to install VariantAnnotation and its dependencies unless they want to import from VCF (there are other import options available).  In any event I now have a published manuscript about polyRAD that points to CRAN as the repository where it is hosted, so changing repositories would probably be confusing for users.

If I look at the checks for the version of polyRAD that is currently on CRAN, the errors about an invalid VCFHeader object only show up for Windows.  However, with the new version that I submitted, CRAN also sends me a Debian check with that same error.  Anyway, I will let them know about Rhtslib.

Best wishes,

Lindsay Clark, Ph.D.
Research Specialist
Department of Crop Sciences
University of Illinois, Urbana-Champaign

-----Original Message-----
From: Pages, Herve <hpages using fredhutch.org> 
Sent: Wednesday, May 1, 2019 11:50 PM
To: Clark, Lindsay V <lvclark using illinois.edu>; bioc-devel using r-project.org
Subject: Re: [Bioc-devel] readVcf problems in package check

Hi Lindsay,

Is this issue specific to the Windows CRAN checks?

This Windows-specific error was also affecting several Bioconductor packages until recently but has been addressed in the latest version of Rhtslib (v 1.15.6). What version of Rhtslib are they using for the CRAN checks? They need to make sure to use the version that belongs to BioC 3.9:

   http://bioconductor.org/packages/3.9/Rhtslib

which is the version that is known to work with R 3.6.0.

BTW have you considered submitting your package to Bioconductor? We typically don't provide support for CRAN packages even if they depend on other Bioconductor packages. In other (cruder) words, by submitting to CRAN, you're on your own.

Cheers,

H.


On 5/1/19 15:07, Clark, Lindsay V wrote:
> I'm the maintainer of polyRAD, which is hosted on CRAN but dependent on several Bioconductor packages including VariantAnnotation.  I'm not sure if my problem is on CRAN's end or Bioconductor's end, but in any event I have not gotten a response from CRAN yet so I thought I would ask here.
>
> I am running into a package check issue which is causing my CRAN submission to be rejected, but which I cannot reproduce.  I have tried fresh installs of R 3.6 and all package dependencies on three different computers and don't see this same error that CRAN reports.  My vignette uses a function that internally calls VariantAnnotation::readVcf, and when the vignette is run, CRAN reports the errors:
>
>
>    Error: processing vignette 'polyRADtutorial.Rmd' failed with diagnostics:
>
>    invalid class "VCFHeader" object: 1: 'info(VCFHeader)' must be a 3 
> column DataFrame with names Number, Type, Description
>
>    invalid class "VCFHeader" object: 2: 'geno(VCFHeader)' must be a 3 
> column DataFrame with names Number, Type, Description
>
> I can run readVcf on the same file myself and see that the header contains info and geno DataFrames that are formatted exactly as described, so I have no idea how that error is occurring.
>
> More confusing (and this makes me suspect that something is wrong with CRAN's check) are the messages:
>
>    [E::bcf_hdr_parse_line] Could not parse the header line: "##contig=<ID=01,assembly=""Msi7"">"
>
>    [W::bcf_hdr_parse] Could not parse header line: 
> ##contig=<ID=01,assembly=""Msi7"">
>
>    [W::bcf_hdr_check_sanity] PL should be declared as Number=G
>
> I found these problems with the file (which is contained in inst/extdata), fixed them, resubmitted the package, yet somehow still get the same message from CRAN's checks.
>
> If any of you want to try performing the checks yourselves, my package is hosted here: https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_lvclark_polyRAD&d=DwICAg&c=eRAMFD45gAfqt84VtBcfhQ&r=BK7q3XeAvimeWdGbWY_wJYbW0WYiZvSXAJJKaaPhzWA&m=k3YdEzkx33T7tU3oJOCTF75TwVTnKpuZZGVEXnho13A&s=5zn32_iC0gM0KtfttAxA53zxyqFL3XG1NTFdrdQGCRw&e=.  The call to VCF2RADdata within the vignette is where these mysterious problems are occurring.
>
> Any help is greatly appreciated.  Perhaps this is just some weird issue with the transition from R 3.5 to 3.6.
>
> Best wishes,
>
> -Lindsay
>
> Lindsay Clark, Ph.D.
> Research Specialist
> Department of Crop Sciences
> University of Illinois, Urbana-Champaign
>
>
> 	[[alternative HTML version deleted]]
>
> _______________________________________________
> Bioc-devel using r-project.org mailing list
> https://urldefense.proofpoint.com/v2/url?u=https-3A__stat.ethz.ch_mail
> man_listinfo_bioc-2Ddevel&d=DwICAg&c=eRAMFD45gAfqt84VtBcfhQ&r=BK7q3XeA
> vimeWdGbWY_wJYbW0WYiZvSXAJJKaaPhzWA&m=k3YdEzkx33T7tU3oJOCTF75TwVTnKpuZ
> ZGVEXnho13A&s=7_wAV-QI5Ukxt5rkf7brSdbRiI60Rj7g80x40jsEY_M&e=

--
Hervé Pagès

Program in Computational Biology
Division of Public Health Sciences
Fred Hutchinson Cancer Research Center
1100 Fairview Ave. N, M1-B514
P.O. Box 19024
Seattle, WA 98109-1024

E-mail: hpages using fredhutch.org
Phone:  (206) 667-5791
Fax:    (206) 667-1319