[Biopython-dev] Thoughts on NN module

[Michiel de Hoon]

I would be in favor of removing Bio.NeuralNetwork. It was useful in the past, but nowadays it seems out of scope for Biopython, and if scikit-learn offers a better solution then we don't need this module in Biopython.
Best,-Michiel 

    On Thursday, June 8, 2017 1:36 AM, Peter Cock <p.j.a.cock at googlemail.com> wrote:
 

 Hi Adil,

So you'd be in favour of us deprecating and later removing
Bio.NeuralNetwork which the original author Brad suggested
on the other thread?

http://mailman.open-bio.org/pipermail/biopython-dev/2017-June/021719.html

Regards,

Peter

On Wed, Jun 7, 2017 at 5:18 PM, Adil Iqbal <aiqbal85 at gmail.com> wrote:
> Hey guys,
>
> I wanted to try and save the machine learning module somehow, but based on
> the varied applications of machine learning in bioinformatics, really the
> best package to use is scikit-learn. It covers everything from Bayesian
> networks, hierarchical clustering, decision trees, etc.
>
> I suppose biopython could offer pre-built, easily modified networks for the
> most common use-cases, that is, protein folding and gene pattern
> recognition. Though we would need to use scikit-learn (or other machine
> learning package) as a dependency.
>
> Another option would be to scrap the machine learning suite and focus more
> on data management.
>
> Best,
> Adil
>
> _______________________________________________
> Biopython-dev mailing list
> Biopython-dev at mailman.open-bio.org
> http://mailman.open-bio.org/mailman/listinfo/biopython-dev
_______________________________________________
Biopython-dev mailing list
Biopython-dev at mailman.open-bio.org
http://mailman.open-bio.org/mailman/listinfo/biopython-dev


   
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.open-bio.org/pipermail/biopython-dev/attachments/20170609/ec734851/attachment.html>