Protocol Online logo
Top : Forum Archives: : Protein and Proteomics

amino acid homology - %identities vs %positive identities (Nov/10/2008 )

Hi,

I have perfomed protein homology (BLAST) search between 2 proteins and I get the following results:

Score = 45.8 bits (107), Expect = 1e-09, Method: Compositional matrix adjust.
Identities = 47/177 (26%), Positives = 62/177 (35%), Gaps = 60/177 (33%)

Do evaluate the % homology, do I look at %idenities or %positives?

Thanks

-Sarwat-

QUOTE (Sarwat @ Nov 11 2008, 11:52 AM)
Hi,

I have perfomed protein homology (BLAST) search between 2 proteins and I get the following results:

Score = 45.8 bits (107), Expect = 1e-09, Method: Compositional matrix adjust.
Identities = 47/177 (26%), Positives = 62/177 (35%), Gaps = 60/177 (33%)

Do evaluate the % homology, do I look at %idenities or %positives?

Thanks


Sorry have to clarify. I meant percent amino acid identity not homology. Mis use of word here.

-Sarwat-

With amino acids, percent similarity is *usually* the best indicator of possible homology.

It depends on a few things -- what matrix is being used to define conservative replacements, where the replacement is (if it's in a catalytic site or other motif, it might not be conservative at all and result in a loss or change in enzymatic activity, whereas if it's in a something like a transmembrane region, it probably won't alter protein function).

-HomeBrew-