[BiO BB] protein identity

Philipp Pagel philipp.pagel at cmp.yale.edu
Mon Oct 22 17:49:43 EDT 2001


On Mon, Oct 22, 2001 at 01:42:08PM -0700, nabula easter wrote:
>   When two proteins are 98-99% identical, what is the
> relationship associated between them. 
>  Can we call those proteins as very close homologues !
> or that 2% difference between then is due to sequence
> errors at Nucleic acid level.
> At nucleic acid level they would be 94% identical and
> we can say a sequnce error!

That depends on how reliable your sequence is. If your sequence data is 
reliable you have a close homologue. What exactly that means depends on the
source of the DNA and functional data:

	- In the same species you might be looking at different isoforms - maybe 
	  with functional differences.
	- If it's from different species it is likely that they are functionally 
	  the same thing.
  - The proteins might also have significantly different functions although 
    you would expect less identity.

So they could be orthologs, paralogs, isoforms or just sequencing errors. Was 
this a hypothetical question or are you talking about something you are working 
on right now? If it's a real-world question you may want to tell us where the
sequence data came from and we can narrow it down some more.

cu and hope it helps


Dr. Philipp Pagel
Department of Cellular and Molecular Physiology         phone: (203) 785-6835	
SHM, B117                                               fax:   (203) 785-4951
Yale University
333 Cedar ST
New Haven, CT 06520

More information about the BBB mailing list