Difference between revisions of "Speed up analysis"
(Forced creation of a TOC - this will improve the formatting on the main page 'Did you know...' section.) |
m (Switch to the {{relax developer link}} template to remove dead Gna! links.) |
||
| Line 2: | Line 2: | ||
== Notes on time consumption == | == Notes on time consumption == | ||
| − | See comments from | + | See comments from {{relax developer link|username=bugman|text=Edward}}. <br> |
[http://article.gmane.org/gmane.science.nmr.relax.user/1468 mailing list correspondance] | [http://article.gmane.org/gmane.science.nmr.relax.user/1468 mailing list correspondance] | ||
Latest revision as of 09:43, 16 October 2020
Contents
Notes on time consumption
See comments from Edward.
mailing list correspondance
Example
A performed 'cpmg fixed' relaxation dispersion analysis, for a dataset with 68 residues.
Analysed with 22 intensity files, with 4 triple replications.
Time consumption was approximately 20 hours.
The analysed models were: R2eff', 'No Rex','LM63','CR72'
What takes time ?
Precision
By default relax uses much higher precision optimisation than most other softwares.
This is based on the philosophy that more accurate results are worth the wait,
especially considering that this time is relatively small in comparison to the measurement and processing time
(and adding the inevitable re-measurements). <
Monte Carlo simulation for accurately determining parameter errors
Lots of Monte Carlo simulations (500 is standard) are used for really accurately determining parameter errors.
For comparison, note that a full a Lipari and Szabo model-free analysis can take between 1 day and 2 weeks to complete.
How to speed up initial phase
For initial analyses where errors are not so important, the number of simulations can be dropped massively to speed things up.
If errors are not important for specific cases, set the number of MC sims to 3, and the analysis will perform much more rapidly.
The result is that the error estimates of the parameters are horrible but, but in some cases, excluding publication, that is not such a problem.
This will not affect model selection either.
