How much homology is needed to function?

[Andrew Henry]

In the referenced article, SHICKLEY at VM.TEMPLE.EDU writes:

>... it all started with examining relatively SMALL areas of homology on
>primary sequences.  The "patches" of homology did show up on helical
>regions which probably need to be aligned for the binding to occur, but
>it was suprising to me that something functional actually popped out of
>the primary sequence homology studies.  My feeling is now that some
>homologies, albiet small or with low overall scores, may have an
>interesting function which may be uncovered with some experimentation.

I think the key phrase here is "patch".  The residues
should be close together in the structure and they may
or may not be contiguous on the sequence.  If you have 
half a dozen residues which are spread out along a piece 
of beta strand, they are not going to be able to all 
bind a small molecule.  Adjacent residues on an alpha 
helix point in different directions.  Working out the
relative positions for residues in an antibody is
possible with reasonable accuracy.  Doing the same for 
sequences of unknown structure and low homology is not
so simple.


		Andrew  Henry

Biochemistry Department			Tel   (0225) 826826 ext 5411
University of Bath			Fax   (0225) 826449
Great Britain.				email  bspahh at gdr.bath.ac.uk