[BioRuby] Blast Format8 hsps' patch

Toshiaki Katayama ktym at hgc.jp
Mon Sep 3 14:11:40 UTC 2007


Hi Raoul,

Thank you for pointing this bug out.

I think I fixed this problem but for double check,
could you personally send me your patch as a separate mail?
This list doesn't allow attachments.

Regards,
Toshiaki

On 2007/08/29, at 18:21, Raoul Jean Pierre Bonnal wrote:

> Hi Guys, 
> I found a bug in the parser. Take this example:
>
>
> contig00002     gi|15902044|ref|NC_003098.1|    99.71   1399    4
> 0       5       1403    414678  416076  0.0     2742
> contig00002     gi|118090026|ref|NC_003028.2|   98.25   858     5
> 1       556     1403    448891  449748  0.0     1592
> contig00003     gi|116515308|ref|NC_008533.1|   99.67   2997    7
> 2       1       2994    423818  426814  0.0     5848
> contig00003     gi|15902044|ref|NC_003098.1|    99.67   2997    7
> 2       1       2994    416288  419284  0.0     5848
> contig00003     gi|118090026|ref|NC_003028.2|   99.60   2997    9
> 2       1       2994    449959  452955  0.0     5832
> contig00004     gi|118090026|ref|NC_003028.2|   98.08   2238    40
> 3       5       2242    453000  455234  0.0     4072
> contig00004     gi|116515308|ref|NC_008533.1|   97.94   2238    43
> 3       5       2242    426859  429093  0.0     4048
> contig00004     gi|15902044|ref|NC_003098.1|    97.94   2238    43
> 3       5       2242    419329  421563  0.0     4048
>
>
> about last contig00003 the parser create an hit with 2 hsps, putting
> togheter results from last contig00003 and first contig00004, this is
> wrong. The code check only if the target is different from the prev one,
> in this case the target is the same but the query is different. 
>
> In attach the patch to solve the problem, I put a chk on the query too.
>
> Actually I don't know if the problem is present with other parsers.
>
> Best regards.
>
> --
> Ra
>
>
> _______________________________________________
> BioRuby mailing list
> BioRuby at lists.open-bio.org
> http://lists.open-bio.org/mailman/listinfo/bioruby




More information about the BioRuby mailing list