>>to models built on low-homology structures.
>
>since i'm currently preparing my bioinformatics lectures for next week's
>teaching, i might as well be a Besserwisser and point out that homology,
>much like pregnancy and death, is a binary concept. i'm sure artem knows
>this and simply mistyped "low sequence identity"
Well, although it is off-topic:
Random House Unabridged Dictionary
Homologous
1. Corresponding or similar in position, value, structure, or function.
So if you insist that homologous is a binary concept then you should be
able to come up with the exact boundary between what's homologous and
what's not.
What is it? 10% sequence identity? Less? More? Because if such a boundary
cannot be defined then everything can be homologous to everything - it's
all in the eye of the beholder. And if so, then the binary concept of
homology is either meaningless or incorrect.
Ergo: arguing about definitions of terms used to describe continua is not
very productive in science (cf. "species", "sea/ocean", "hill/mountain").
Dima
|