Hi, Thanks a lot for many clarifications and useful comments in your answer. Much appreciated. Most important for me, the value.write() command solved the main issues, now I easily get all s2, te, and chi2.
One, quite small, question mark remains. How do I get an estimate of the global correlation time (tm)? I use one magnetic field only, I have read in many comments on the mail list that one field is suboptimal, for many reasons, I’m aware of this, but since my protein (in)stability doesn’t admit more NMR-time, one field I hope is better than no field! So based on my analysis, Protein ca 400 residues one field 800 MHz, T1-, T2- and NOE-data for most residues is there any neat way to get an estimate of the global correlation time (or diffusion tensor) from relax? Or is this only possible using two or more fields? (I could get a rough estimate of tm from other software, e.g. David Fushman’s Matlab-based ROTDIF.) I apologise if the answer to my question is already clearly posted, I’ve searched through quite many posts on the mail list to try to find answers or protocols to get a rough estimate of tm, but I haven’t found any. Best regards Johan Wallerstein > On 26 Nov 2021, at 12:26, Edward d'Auvergne <[email protected]> wrote: > > On Fri, 26 Nov 2021 at 12:14, Johan Wallerstein <[email protected]> > wrote: >> >> Hi Edward, >> >> Thanks for all efforts put into developing the relax-software. > > Hi Johan, > > You're welcome! And welcome to the relax mailing lists! > > >> I'm using the model-free script ‘single_model.py', and get as output an >> XML-file called ‘results’ ca 5 Mb. >> >> How do you suggest the user to best deal with that file? > > That file contains all data you would ever require. It is probably > best to use relax to extract what you need - just load that file into > a new instance of relax and use the appropriate relax users functions > to output text lists or graphs of the data. > > >> More specifically what confuses me is that I get many lists of ‘s2_sim’ (see >> below), are we supposed to collect all of these lists and compute mean and >> standard deviation? > > These are the values fitted for each of the Monte Carlo simulation > data sets. The relax script you used should already have calculated > the standard deviation for this and placed it into the *_err data > structures. Note that the average value of fitted Monte Carlo > simulations has no practical or statistical uses so we never calculate > it. > > >> Or do you suggest me to retrieve the s2-data from the log-file? >> (using command 'relax —log log.txt single_model.py’) > > It's best to simply load the file back into relax and use user > functions such as value.write(), grace.write(), pymol.macro_write(), > molmol.macro_write(), etc to output the data in a way you can use it - > text lists with values and errors, 2D graphs with error bars, or > values represented via a 3D structure. > > >> before when I've used the ‘dauvergne_protocol.py’ I got all output data in >> the folder ‘final', and as text files. > > There should also be Grace graph *.agr files in there, as well as > PyMOL macros. I hope this information helps. > > Regards, > > Edward _______________________________________________ nmr-relax-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/nmr-relax-users
