Biomart: Biomart Or Perl Api Or Xml Query

Question

[Tom]

Hi all,

I have used biomart quite often in the past but mainly using a Perl API or R biomaRt package. I realize there is a shift to biomart 0.8 and that both methods that I use are not yet compatible. My two short questions are;

1) Does it mean that using Perl or R I am not able to access the newest data? I am confused what does the 0/7 to 0/8 transition means to me.

2) If I am building software that extracts automatically data from biomart what would be the best and most future proof method? I really like perl and R methods but if they can be obsolete in the future is it better to build XML queries? My queries can extract huge amount of data so I would like to avoid server time outs etc.

Thanks, Tom

Answer 1

[Neilfws]

This was discussed on the BioMart users mailing list:

Q: Just curious, but will the perl API be supported long-term?

A: No, at this time, we don't have any plan for supporting Perl API in the BioMart 0.8. It is encouraged to use REST API for remote access BioMart data sources.

Perl API needs to be rewritten in order to work with BioMart 0.8 (and later versions). Basically, it would be a wrapper of the REST API, just like how 'biomaRt' R package works.

This suggests to me that (1) the current Perl wrapper will be discontinued; (2) someone may write a replacement and (3) Bioconductor/biomaRt should continue to work. Development is very active on most Bioconductor projects, so I would expect a lot of effort in that area.

In terms of future proofing: sounds like the best option is to install a local test instance, get to grips with the REST API and write some code to deal with REST queries and output. It's "just" a case of constructing a URL and parsing whatever is returned (XML, JSON), which is pretty standard and generic in most languages.

[SES]

+1 for finding this information, and for the thorough response. I've been wondering what was happening with the Perl API and how we should access BioMart in the future.

Answer 2

[Arek]

Hi Tom, we are running two versions of BioPortal in parallel (0.7 and 0.8). They talk to the same databases therefore the they have always the same version of the data

a