Re: diffusion tensorRe: diffusion tensorHi Angelo, I'm not sure if you have seen this as you may not be subscribed to the relax-announce mailing list (https://mail.gna.org/listinfo/relax-announce), but a new version of relax has been released fixing the bugs you have encountered. The following link describes the different ways in which you can find out about new relax releases: http://article.gmane.org/gmane.science.nmr.relax.user/1413 Regards, Edward On 13 March 2013 18:15, Edward d'Auvergne <edward@xxxxxxxxxxxxx> wrote:
Hi, Have you tried Sebastien Morin's consistency testing analysis which is in relax (http://www.nmr-relax.com/manual/Consistency_testing.html)? This might tell you if there are problems with your relaxation data. I would assume that you have used proper per-experiment temperature calibration using MeOH/ethylene glycol and single-scan interleaving for temperature control: http://www.nmr-relax.com/manual/Temperature_control_calibration.html And that you have used the appropriate spectral processing and peak height extraction: http://www.nmr-relax.com/manual/From_spectra_peak_intensities_relaxation_rates.html These links are from the HTML version of the relax manual (http://www.nmr-relax.com/manual/index.html). Note that the error analysis is incredibly important - I always say that accuracy in the errors is just as important, or maybe even more important, than the data itself. If the errors are wrong, then the results from any type of modelling analysis will be meaningless (there are shelves in maths libraries dedicated to this large field of mathematics: http://en.wikipedia.org/wiki/Mathematical_model) . Assuming you have done all of this, then there could be other issues affecting the analysis. Did you collect all data on a single sample and, if not, did you make sure the protein concentration was identical in each sample? Partial dimerisation, even non-specific and at very low percentages, could significantly affect the analysis. Even more so if slightly different sample concentrations are used. Did you use all of the global diffusion models in the model-free analysis? Are the AIC values significantly different between the models? Also, did you use relax to calculate the R1 and R2 relaxation rates and the NOE? Note that almost no protein will tumble as a perfect sphere. You can see this with the inertia tensor, and almost no model-free analysis using today's methodology will select the spherical tensor. One part that you don't see, which is a very significant part of the diffusion tensor, is the water shell. This can have different thicknesses around your system depending upon surface hydrophobicity/hydrophilicity, the presence of charges, bound metals (even a loose attraction), and loop and other internal motions. So even a perfect mathematical sphere with different localised surface properties will not diffuse as a sphere. Anyway, I hope some of this information helps. Regards, Edward On 13 March 2013 17:31, Angelo Miguel Figueiredo <am.figueiredo@xxxxxxxxxx> wrote:Sorry once again. In fact relax has chosen the 'local_tm" however I don't get any diffusion tensor even selecting on the user function menu the diffusion_tensor.display says: "RelaxError: No diffusion tensor data exists" I think this might be normal for the local_tm model once it assumes the molecule doesn't diffuse as a globular protein and somehow none of the models fits the data?! However, this result is quite strange since my protein is quite spherical and the inertia tensors are 1: 0.8: 0.7 Any ideas? many thanks, Angelo On 13 Mar 2013, at 13:07, Edward d'Auvergne wrote:Hi Angelo, Welcome to the relax mailing lists! The problem you are seeing is a clear bug. Would you be able to submit a bug report using the link https://gna.org/bugs/?func=additem&group=relax ? Thanks. Although bug #12408 (https://gna.org/bugs/?12408) appears to be similar, this is only because both cases use the special RelaxNoTensorError error object. For the bug report, would you be able to include the output of: $ relax --info Also if you could look at the logs and say which global diffusion model was selected, that would be useful. And have you used a 3D structure file for the analysis? Note that this failure has occurred at the very end of the analysis. The call to the structure.create_diff_tensor_pdb user function is the very last thing that this model-free auto-analysis does (you can see that here if you are interested, http://www.nmr-relax.com/api/auto_analyses.dauvergne_protocol-pysrc.html#dAuvergne_protocol.execute). Therefore this error is not too important for your analysis. It would be good to have the bug report though so that I can come up with a solution. Cheers, Edward On 13 March 2013 13:43, Angelo Miguel Figueiredo <am.figueiredo@xxxxxxxxxx> wrote:Dear relax community, I am a new user that recently installed relax 2.2.2. After been playing around with it I spot in the late stage of my final calculation the following message in attach. Does anybody has a clue or advice what could be wrong with it. I spot a similar error in earlier relax versions (bug #12408) but I am not sure if it is the same issue, apparently relax can't read/load my diffusion tensor. Any help will be appreciated. Many thanks, Angelo relax> structure.create_diff_tensor_pdb(scale=1.8e-06, file='tensor.pdb', dir='/home/final', force=True) Exception raised in thread. Traceback (most recent call last): File "/home/angelo/Software/relax-2.2.2/gui/analyses/execute.py", line 87, in run self.run_analysis() File "/home/angelo/Software/relax-2.2.2/gui/analyses/auto_model_free.py", line 808, in run_analysis dauvergne_protocol.dAuvergne_protocol(pipe_name=self.data.pipe_name, pipe_bundle=self.data.pipe_bundle, results_dir=self.data.save_dir, diff_model=self.data.global_models, mf_models=self.data.mf_models, local_tm_models=self.data.local_tm_models, grid_inc=self.data.inc, diff_tensor_grid_inc=self.data.diff_tensor_grid_inc, mc_sim_num=self.data.mc_sim_num, max_iter=self.data.max_iter, conv_loop=self.data.conv_loop) File "/home/angelo/Software/relax-2.2.2/auto_analyses/dauvergne_protocol.py", line 234, in __init__ self.execute() File "/home/angelo/Software/relax-2.2.2/auto_analyses/dauvergne_protocol.py", line 745, in execute self.write_results() File "/home/angelo/Software/relax-2.2.2/auto_analyses/dauvergne_protocol.py", line 937, in write_results self.interpreter.structure.create_diff_tensor_pdb(file="tensor.pdb", dir=dir, force=True) File "/home/angelo/Software/relax-2.2.2/prompt/uf_objects.py", line 219, in __call__ self._backend(*new_args, **uf_kargs) File "/home/angelo/Software/relax-2.2.2/generic_fns/structure/geometric.py", line 449, in create_diff_tensor_pdb raise RelaxNoTensorError('diffusion') RelaxNoTensorError: RelaxError: No diffusion tensor data exists. _______________________________________________ relax (http://www.nmr-relax.com) This is the relax-users mailing list relax-users@xxxxxxx To unsubscribe from this list, get a password reminder, or change your subscription options, visit the list information page at https://mail.gna.org/listinfo/relax-users_______________________________________________ relax (http://www.nmr-relax.com) This is the relax-users mailing list relax-users@xxxxxxx To unsubscribe from this list, get a password reminder, or change your subscription options, visit the list information page at https://mail.gna.org/listinfo/relax-users