correct use of homology-identity-similarity ?

Vladimir Svetlov svetlov at
Mon Nov 30 15:59:36 EST 1998

In article
<Pine.A41.4.05.9811292315390.42260-100000 at>,
"Carsten Hohoff" <hohoff at> wrote:

> Where can I find the definition of 'homology', 'similarity' and 'identity'
> in sequence alignments ?
> We just had a discussion in the lab about this topic - I thought, 'identical' 
> amino acid means the same aa, while  'similar' means a related (e.g., basic, 
> acid, hydrophobic etc.) and 'homolog' is just a rough description of a
> relationship based on sequence alignment...
> Any hints that could help clarifying these terms are highly appreciated

Firstly, if the terms originated with particular computer-assisted
alignment/philogeny program, consult the manual/original paper/designers
for definitions - you'd be surprised what a range of meanings the term
"similar" can take. Often similar means "within so many units from..."
according to a particular similarity scoring table (like BLOSUM or PAM). In
the literature there is a great discordance as to what amino acids are
"similar" and this is reflected in the assessment of similarity by various
programs out there.
Secondly, according to conventional wisdom in a given field the word
"homolog" can mean different things - from simple statement of sequence
similarity to a documented evolutionary relationship with all that
homotopy, homonomy and stuff.
So it depends on how rigorous you wanna be in using this term.


Vladimir Svetlov, PhD
McArdle Lab for Cancer Res.
1400 University Ave.
Madison, WI 53706

More information about the Methods mailing list