[BiO BB] Single-domain proteins

[jfreeman]

Hi Dr. Suraj, et al.

On the bioinformatics end of things:

If a protein has two or more domains, but it is only annotated with a
single domains function, and a simple homology from that proteins domain
to another protein will give transitive false annotation assignments.

For example:

1. A--
2. A--B
3. B--C
4. C--

Given that: 1, 2, 3, and 4 are different proteins.

and domains A, B, C, and D have different functions, but you only know A
and C's functions, what you annotate protein 2 and 3 depends on what you
start with 1 or 4.  If you use simple homology to assign a function to 2
from 1, and then use this functional assignment on 2 to annotate by
simple homology protein 3, the annotation for 3 will be function A and
be completely false.

Finding single domain proteins, or important motifs, therefore becomes a
very important annotation tool.

For more information see:

http://bmerc-www.bu.edu/bmerc_info/temple.html
http://bmerc-www.bu.edu/bioinformatics/profile_list.shtml
http://www.expasy.ch/prosite/
http://pfam.wustl.edu/
http://www.geneontology.org/

Happy Hunting,

Jim Freeman


Peri Suraj wrote:
> 
> hi group,
>   what are the advantages of single-domain proteins ?
> 
> thanks
> SP
> 
> =====
> PIL/CEBI/SDU/DK
> 
> __________________________________________________
> Do You Yahoo!?
> NEW from Yahoo! GeoCities - quick and easy web site hosting, just $8.95/month.
> http://geocities.yahoo.com/ps/info1
> 
> _______________________________________________
> BiO_Bulletin_Board maillist  -  BiO_Bulletin_Board at bioinformatics.org
> http://bioinformatics.org/mailman/listinfo/bio_bulletin_board