There are three general approaches to molecular replacement:
1. rotation-translation algorithms: these divide the 6-dimensional
problem of placing a replacement model in the ASU into two
sequential 3-parameter searches, which, if it works, converges more
quickly than a 6-dimensional search. Examples include Amore, MOLREP,
Phaser, etc.
2. genetic algorithms: these use an evolutionary search model to more
quickly converge a 6-dimensional search. Example: EPMR.
3. brute-force algorithms: these use brute-force stochastic methods to
attempt to find the global best fit to a 6-dimensional search for
the best placement of the model in the ASU. Example: Queen of Spades.
Each of these methods will have their strengths and weaknesses. The
brute-force algorithms nearly always work but may take a very, very long
time. For routine problems the rotation-translation or genetic
algorithms are quite efficient. (Some are faster than others.) When
searching for multiple models in the ASU, some algorithms may work
better than others for a particular case. In my experience EPMR and
Phaser seem to be very good at accurately placing multiple models within
an ASU in a reasonable amount of time, but all these programs have their
quirks when trying to solve specific problems, or when working with
low-homology search models. So it's good to have a variety of approaches
to choose from.
Cheers,
_______________________________________
Roger S. Rowlett
Gordon & Dorothy Kline Professor
Department of Chemistry
Colgate University
13 Oak Drive
Hamilton, NY 13346
tel: (315)-228-7245
ofc: (315)-228-7395
fax: (315)-228-7935
email: rrowl...@colgate.edu
On 4/22/2013 3:05 AM, LISA wrote:
Hi all,
There are so many software for MR, such as phaser, balbe,molrep, and
amore. What is difference between them? Which one is powerful?
Please give some comments for these software?Thank you.
Sincerely,
lisa
