[Biopython-dev] Thoughts on API changes for multiple file versions of Affy CEL format

[Adam Kurkiewicz]

A few remarks regarding this pull request:

Effort has been made for the changes to the API to be backwards
compatible: any code which worked with the previous API should work
the same way with the new API.

It would be helpful if any users of the Bio.Affy module tried to run
their code against the version of the module from this pull request
and check if they still get the same behaviour.

On a somewhat broader note, I'd be interested in how the users of the
Affy package *actually* use it. At the moment, the package provides
only stage 1 of microarray analysis (reading raw data).

At least 3 stages are required for end-to-end data analysis:
  1. reading raw data (already provided).
  2. normalisation & background correction (including aggregation of
probes into probesets or transcription clusters).
  3. gene discovery (e. g. linear models, like limma), or any other
form of data analysis.

Is there any code in Biopython which does stages 2 and 3? Have people
written their own code for stages 2 and 3 or is everybody using R?
Would there be any broader interest in creating common tools for stage
2 an 3?

Adam

On 29 November 2016 at 11:48, Peter Cock <p.j.a.cock at googlemail.com> wrote:
> RE: https://github.com/biopython/biopython/pull/1011
>
> In this pull request to add support for the (binary) CEL v4 format,
> changes to the existing API are being suggested (for the text v3
> format only).
>
> I'd like a few more eyes on this please (ideally on the pull request),
> especially anyone who is using the Bio.Affy code.
>
> Thanks,
>
> Peter
> _______________________________________________
> Biopython-dev mailing list
> Biopython-dev at mailman.open-bio.org
> http://mailman.open-bio.org/mailman/listinfo/biopython-dev