--
- Dr. Azat Badretdinov
- The Rockefeller Univ, Box 270
- 1230 York Ave, New York NY 10021, USA
- Phone: (212) 327 7206
- Fax: (212) 327 7540
- E-mail:
- WWW/URL: http://salilab.org/~azat
--- Begin Message ---
To: Nick Glover <>
Subject: Re: alignment queries..
From: Azat Badretdinov <>
Date: Tue, 10 Feb 1998 10:13:44 -0500
Organization: Rockefeller University
Dear Nick!
replying...
> Question 1:
>
> What is considered to be the minimum length of a sequence motif
> necessary to derive meaningful constraints from the alignment to use in
> modeling.. one, two, three, or more?
>
> i.e.
>
> template -K- -YI- -EEE- -YSIG- -DIDAP-
> target -x- -xx- -xxx- -xxxx- -xxxxx-
>
> The Modeller routines tend to pick out longer motifs than Meme,
> Matchbox, PIMA, BLOCKS, etc. Smallest motif I have had picked out for
> my sequences using these alternative methods is three a.a.'s...
>
> what would be the effect of only matching two, or even just one residue
> on the modeling?
Usually more than that (dozens if you want just to detect reliable
similarity, and even more if you want a real model).
It is good to have at least 35-40% sequence identity to build a model.
Sometimes even 30% is Ok.
>
> which brings me to my next query..
>
> Question 2:
>
> how rigorous must one be with the alignment?
>
> given an template of, say, IPSDC would it be okay to match VPGCC, etc..
> trying, in general, to keep acidic matched to acidic, basic to basic,
> rings to rings and hydrophobic to hydrophobic, bulky to small, as shown
> in (A)..
>
> is this the right method - be less worried about the *exact* nature of
> the match, as long as they are physically/chemically similar.. else, is
> it best to mix and match among templates, as illustrated in (B)
>
> (A) is this acceptable?
>
> template 1 LPSDC
> template 2 VQSAC
> template 3 IPSEC
> template 4 VPGCC
> target IPSDC
>
> (B) or should I break it up like this?
>
> template 1 L-P-S-D-C
> template 2 V--QS--AC
> template 3 -IP-S--EC
> template 4 V-P--G-CC
> target -IP-S-D-C
>
> I believe that this is the critical step in the whole procedure, so
> your comments would be greatly appreciated!
I this case I guess you can trust multiple alignment program which will
probably produce case A.
--
- Dr. Azat Badretdinov
- The Rockefeller Univ, Box 270
- 1230 York Ave, New York NY 10021, USA
- Phone: (212) 327 7206
- Fax: (212) 327 7540
- E-mail:
- WWW/URL: http://salilab.org/~azat