[Bioperl-l] free energy of two RNA

[Barry Moore]

Vesselin-

I'll preface these comments by saying that this is not my field, so I 
only know enough about it to be dangerous.  John Santa Lucia has a web 
server running a program called Hyther that will accomplish your task 
using a nearest neighbor model which I believe is currently the most 
accurate way to predict thermodynamics from sequence  
(http://ozone2.chem.wayne.edu/Hyther/hytherm1main.html).  Unfortunately, 
the code is proprietary, so you can't download the program and run it 
locally.  If you were hoping to do lots of these calculations, you've 
got a couple of options.  The Hyther web server uses form posts for 
submission.  One option would be two write some code using Perl's LWP 
module (http://search.cpan.org/~gaas/libwww-perl-5.79/lib/LWP.pm) that 
would send off lots hits to the web server.  If you want to do it in a 
more robust way, you could write some code of your own to do this.  The 
BioPerl module Bio::SeqFeature::Primer.pm 
(http://cvs.bioperl.org/cgi-bin/viewcvs/viewcvs.cgi/bioperl-live/Bio/SeqFeature/Primer.pm?cvsroot=bioperl) 
has a Tm method that uses nearest neighbor parameters to calculate 
melting temperature predictions.  The current code calculates enthalpy 
and entropy for DNA from a single input sequence.  To modify it to you 
needs would involve changing the nearest neighbor parameters to the 
correct ones for RNA, and adding the RNA free energy parameters.  There 
are a couple of things to keep in mind if you do this.  One is that the 
way the method is currently written, it only allows for on strand of 
nucleic acid as input.  It sounds like you want to do your calculations 
on two strands that are probably not completely complimentary.  If that 
is the case, then you've got a bit  more programing to do.  You'll have 
to add in a more complete set of nearest neighbor parameters for 
mismatches, and you'll have to add code to detect and score those 
mismatches in your input sequences.  I'm pretty sure that Santa Lucia 
has all those parameters publicly available (he gave a talk here a 
couple days ago, and said as much) although I couldn't find them on his 
web site.  One final caution is that the parameters currently available 
are for 1M Na+ concentrations.  He has published corrections to be added 
for other concentrations of monovalent ions, but the corrections for 
divalents like Mg2+ are again - proprietary.  John gets poor marks for 
behavior - does not share code with others, but the web server does do 
Mg corrections.  Would be interested in hearing about it if you find 
other/better solutions.

Barry


baev at imbb.forth.gr wrote:

> I have to calculate the free energy of binding between two RNA 
> molecules (not
> for each individual RNA). But these program allowed only one 
> sequence?! not
> two?
>
>
>
>
>
> _______________________________________________
> Bioperl-l mailing list
> Bioperl-l at portal.open-bio.org
> http://portal.open-bio.org/mailman/listinfo/bioperl-l


-- 
Barry Moore
Dept. of Human Genetics
University of Utah
Salt Lake City, UT