[Bioperl-l] SeqIO out of memory

Marc Logghe Marc.Logghe at devgen.com
Mon Mar 3 19:16:27 EST 2003


I have found it also a lot of times for the annotation of S-S bridges in
protein sequences.
Anyhow, the bond bug (the name is Bond, Bug Bond ;-) was reported by Aaron a
while ago: http://bugzilla.bioperl.org/show_bug.cgi?id=1371
Cheers,
Marc

> -----Original Message-----
> From: Brian Osborne [mailto:brian_osborne at cognia.com]
> Sent: maandag 3 maart 2003 18:25
> To: Ewan Birney
> Cc: Hilmar Lapp; Bioperl
> Subject: RE: [Bioperl-l] SeqIO out of memory
> 
> 
> Ewan,
> 
> Here's the relevant section:
> 
>      misc_feature    1553..1564
>                      /gene="aap-1"
>                      /note="Region: [PSORT] ER membrane domain: KSIP"
>      misc_feature    bond(174,175)
>                      /gene="aap-1"
>                      /note="Intron length 47 bp, type gt_ag"
>      exon            175..434
>                      /gene="aap-1"
>                      /note="Exon 2 length 260 bp"
>      misc_feature    bond(434,435)
>                      /gene="aap-1"
>                      /note="Intron length 181 bp, type gt_ag"
>      exon            435..700
>                      /gene="aap-1"
>                      /note="Exon 3 length 266 bp"
>      misc_feature    bond(700,701)
>                      /gene="aap-1"
>                      /note="Intron length 227 bp, type gt_ag"
>      exon            701..1132
>                      /gene="aap-1"
>                      /note="Exon 4 length 432 bp"
>      misc_feature    bond(1132,1133)
>                      /gene="aap-1"
>                      /note="Intron length 80 bp, type gt_ag"
>      exon            1133..1802
>                      /gene="aap-1"
>                      /note="Exon 5 length 670 bp"
>      3'UTR           1571..1802
>                      /gene="aap-1"
>                      /evidence=experimental
>      polyA_signal    1789..1794
>                      /gene="aap-1"
>                      /note="variant  attaaa"
> BASE COUNT      560 a    368 c    355 g    519 t
> ORIGIN
>         1 aatgagcaca acacctggaa ctcctcatgg agtcactcat 
> agtcttatgg aacaaggatg
> 
> 
> Yes, it should not be there but there it is. I'm guessing 
> that the thought
> is that "bond" is useful for mRNAs, indicating the positions 
> of the exons on
> the sequence. In the RefSeq file I have there are >100,000 
> entries and some
> 7,000 misc_feature's with "bond", all of them <num>,<num>, 
> not surprisingly.
> The documentation you've provided shows that the general format
> (<num>,<num>) is a valid one, whether to handle "bond" is another
> question....
> 
> Brian O.

> 


More information about the Bioperl-l mailing list