Comments on: Obtaining overrepresented motifs in DNA sequences, part 11 http://python.genedrift.org/2008/08/12/obtaining-overrepresented-motifs-in-dna-sequences-part-11/ a step-by-step guide to create Python applications in bioinformatics Mon, 22 Feb 2010 18:22:18 +0000 http://wordpress.org/?v=3.0-alpha hourly 1 By: Andrew Dalke http://python.genedrift.org/2008/08/12/obtaining-overrepresented-motifs-in-dna-sequences-part-11/comment-page-1/#comment-17632 Andrew Dalke Fri, 15 Aug 2008 06:30:02 +0000 http://python.genedrift.org/2008/08/12/obtaining-overrepresented-motifs-in-dna-sequences-part-11/#comment-17632 For those coming across this blog via a search, check out my comments in part 12 of this series. The "binom" implementation here is really slow, as Paulo's benchmark in #12 shows. That implementation is decidedly not clever. A faster implementation gives results in much shorter time. For the benchmark case, about 3,000 times faster. For those coming across this blog via a search, check out my comments in part 12 of this series. The “binom” implementation here is really slow, as Paulo’s benchmark in #12 shows. That implementation is decidedly not clever.

A faster implementation gives results in much shorter time. For the benchmark case, about 3,000 times faster.

]]> By: cariaso http://python.genedrift.org/2008/08/12/obtaining-overrepresented-motifs-in-dna-sequences-part-11/comment-page-1/#comment-17590 cariaso Thu, 14 Aug 2008 15:31:28 +0000 http://python.genedrift.org/2008/08/12/obtaining-overrepresented-motifs-in-dna-sequences-part-11/#comment-17590 Paulo, hold your ground. Numpy may yet prove useful, but for now it is premature optimization. Your audience will get more value from watching a small app develop and be benchmarked. Already I think the 500x performance difference between the numerically clever binomial approach vs 3 for loops is a good lesson. Keep it up. Paulo, hold your ground. Numpy may yet prove useful, but for now it is premature optimization. Your audience will get more value from watching a small app develop and be benchmarked. Already I think the 500x performance difference between the numerically clever binomial approach vs 3 for loops is a good lesson. Keep it up.

]]> By: Paulo Nuin http://python.genedrift.org/2008/08/12/obtaining-overrepresented-motifs-in-dna-sequences-part-11/comment-page-1/#comment-17588 Paulo Nuin Thu, 14 Aug 2008 12:59:38 +0000 http://python.genedrift.org/2008/08/12/obtaining-overrepresented-motifs-in-dna-sequences-part-11/#comment-17588 I guess in the end the scripts will evolve and I'll include a scientific package. Also the scripts are very simple and straightforward, that sometimes big package might get in the way. I guess in the end the scripts will evolve and I’ll include a scientific package. Also the scripts are very simple and straightforward, that sometimes big package might get in the way.

]]> By: baoilleach http://python.genedrift.org/2008/08/12/obtaining-overrepresented-motifs-in-dna-sequences-part-11/comment-page-1/#comment-17586 baoilleach Thu, 14 Aug 2008 12:09:10 +0000 http://python.genedrift.org/2008/08/12/obtaining-overrepresented-motifs-in-dna-sequences-part-11/#comment-17586 Um...that's an interesting approach, although not one I would recommend in practice. Perhaps at the end, you could present a solution that uses one of the main Python scientific libraries. (This may also yield performance improvements if you feel this is important.) Um…that’s an interesting approach, although not one I would recommend in practice. Perhaps at the end, you could present a solution that uses one of the main Python scientific libraries. (This may also yield performance improvements if you feel this is important.)

]]> By: Paulo Nuin http://python.genedrift.org/2008/08/12/obtaining-overrepresented-motifs-in-dna-sequences-part-11/comment-page-1/#comment-17527 Paulo Nuin Wed, 13 Aug 2008 11:49:14 +0000 http://python.genedrift.org/2008/08/12/obtaining-overrepresented-motifs-in-dna-sequences-part-11/#comment-17527 Yes, there is. I am trying to create the scripts without many imports, using mostly Python's included libraries. Yes, there is. I am trying to create the scripts without many imports, using mostly Python’s included libraries.

]]> By: baoilleach http://python.genedrift.org/2008/08/12/obtaining-overrepresented-motifs-in-dna-sequences-part-11/comment-page-1/#comment-17517 baoilleach Wed, 13 Aug 2008 09:09:54 +0000 http://python.genedrift.org/2008/08/12/obtaining-overrepresented-motifs-in-dna-sequences-part-11/#comment-17517 Is there a reason you're not using numpy, scipy or rpy for maths? Is there a reason you’re not using numpy, scipy or rpy for maths?

]]>