Hi Ed, Ok, that's what I thought after some reflexion... I wanted to ask since I was aware that this could cause chi2 values or model selection to vary, etc... Thus, relax's behaviour is the same as in ModelFree for what concerns which Rex is reported. Thanks Ed ! Séb :) Edward d'Auvergne wrote: Hi, In relax, the Rex reported will be that of the first input field strength. The frequency of the Rex value is documented in the header line of the results file. This is a deliberate design to increase the flexibility of relax - Rex is automatically handled for you. If you would like to convert to a different field strength, just divide by the proton frequency in Hz, squared, then multiply by the new frequency squared. relax could be modified to report the value for the highest field, this will of course also be fully automatic from the user's perspective, but the current design allows an advanced user to choose which frequency the Rex is reported for. Regards, Edward On Feb 18, 2008 4:57 AM, Sebastien Morin <sebastien.morin.1@xxxxxxxxx> wrote:Hi, I have a question concerning Rex and its magnetic field dependence vs how it is treated in relax. Rex, the contribution from us-ms motions on the transversal relaxation rate (R2), is scaled quadratically with the magnetic field. When doing an analysis with relax, one gets a value for Rex. When working with data from multiple magnetic fields, I realized that the dataset to which this Rex is associated changes as a function of the order with which datasets are input into relax. In fact, the Rex value is associated with the first dataset input. Here are some results I had with the full_analysis.py script modified to minimize only tm3 (for a rapid test) : ================================================================ Input_order Association Rex__residue_25 =========== =========== ======================= 500-600-800 500 7.0885287044805514 500-800-600 500 7.0885287044182812 600-800-500 600 2.7486344907905164e-13 600-500-800 600 2.7486344897092221e-13 800-500-600 800 -2.1641012543284559e-17 800-600-500 800 -5.3169783493616144e-21 ================================================================ As you can see, the Rex value changes with the input order. Most change comes from the first dataset input, but there is also a small influence from the subsequent datasets input... Is this normal ? Would it be better to always input datasets so Rex gets associated with the highest magnetic field ? With the lowest ? Thanks for clarifying this issue ! Cheers, Séb -- Sebastien Morin Etudiant au PhD en biochimie Laboratoire de resonance magnetique nucleaire Dr Stephane Gagne CREFSIP (Universite Laval, Quebec, CANADA) 1-418-656-2131 #4530 _______________________________________________ relax (http://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 -- Sebastien Morin Etudiant au PhD en biochimie Laboratoire de resonance magnetique nucleaire Dr Stephane Gagne CREFSIP (Universite Laval, Quebec, CANADA) 1-418-656-2131 #4530 |