 |
 |
Hi, I've always been a little confused as to the point of LabelledMolecule if it required a different name for each labelled incarnation. I'd always thought that LabelledMolecule should be a subset of a particular Molecule that takes into account the isotopic labelling present in the molecule. It therefore seemed to make sense to me for a molecule to be able to have a number of differently labelled possibilities. So I guess I'd thought that Molecule should represent the chemistry of your molecular entity, and that this can be present with different isotopic compositions. Therefore LabelledMolecule should store the different isotopic compositions of a single chemical under the same name as the Molecule. Hope I made clear my confusions. Andy Rasmus Fogh wrote: > Dear Gary, > > This is indeed messy stuff. And something will have to be changed, it > looks like. > > As the data model stands, there can be only one LabelledMolecule for > each Molecule. That is correct - and code that creates more than one > is therefore wrong. > > The correct way to work according to the model would be to have a > separate MolLabel for every molecular labelling, and a separate > LabeledMixture for every distinct sample composition. > > It looks like the ECI is not using the model correctly, as it stands. > The solution could be either to change ECI (and improve the Model > documentation), or to change the model. Or maybe both, because if it > is too hard to do the things you need with the current model, we will > have to look at what we can do. > > We are pretty busy, as you know, but I shall have to look at this with > Pieter and Aleks. > > Is there anyone else who has something to contribute and would like to > be included in the discussion? > > Yours, > > Rasmus > > --------------------------------------------------------------------------- > > Dr. Rasmus H. Fogh Email: [log in to unmask] > Dept. of Biochemistry, University of Cambridge, > 80 Tennis Court Road, Cambridge CB2 1GA, UK. FAX (01223)766002 > > On Tue, 20 Nov 2012, Gary Thompson wrote: > >> Hi wayne >> >> sorry more questions >> >> I am still slightly confused as to how this works... and also >> (possibly wrongly) forsee some possible problems >> >> 0. the labeledMolecules appear to have been created in response to >> creating two samples and their sample components and labelling in the >> eci samples page. I created two different samples with different >> labeling patterns ( named methanol_CN and methanol_N 100% CN and 100% >> N labelled respectivley). >> >>> From the documentation (which is pretty sparse) >> >> 1. A labeledMoleculed represents a labelled version of a molecule >> with a number of labelling patterns which are defined via molLabels >> and children, with there being one molLabel for each labelling >> pattern i.e uniform 13c vs uniform 15N 13c vs labelled 13C at residue >> 1 ca only >> >> 2. a labelledMolecule has a single element as its full key (which >> must be globally unique) and this is its name. This should match the >> name of the molecule the labeledMolecule shadows so there can only be >> one labeledMolecule for each Molecule >> >> 3. i have two labeled molecules (courtesy of eci when I created two >> samples) with two names methanol_N and methanol_CN which don't match >> my molecule >> >> 4. the suggestion is to rename one of them (say the one named >> p7:methanol_CN) to match my molecule. I presume I would then have to >> delete the other labeledMolecule (say the one with the name >> p7:methanol_N )? However then I have lost all the labelling >> information that goes with the other labeledMolecule (ie >> p7:methanol_N) and eci no longer knows about my N labelled sample...) >> >> 5. will this cure cause eci problems (i.e. will there be dangling >> references to the second labelled molecule left over ? ) for example >> the experiments panel contains a sample column with popups that match >> the names of the methanol_N and methanol_CN samples I created. Do >> these refer to samples and if so how do the samples relate to the >> labeledMolecules... >> >> my results from digging around are that samples refere to >> labeledMixtures via the following path, so if I can transfer the >> labeled micture over everything should be alright... >> >> top.project.sortedNmrProjects()[0].sortedExperiments()[<a >> number>].sample.sortedSampleComponents()[0].refComponent.labeledMixture.labeledMolecule >> >> >> sorry clearly plenty of confusion & worries on my part here >> >> regards >> gary >> _________________ >> From: CcpNmr software mailing list [[log in to unmask]] On Behalf >> Of Wayne Boucher [[log in to unmask]] >> Sent: 17 November 2012 00:28 >> To: [log in to unmask] >> Subject: Re: [CCPNMR] problems with data shifter >> >> >> >> Hello, >> >> This indeed seems to be sort-of and ECI bug. The reason I say sort-f is >> that it seems a bit daft to me that the LabeledMolecule constructor >> allows >> the name not to be that of a Molecule. (The constructor checks validity >> but not with "complete=True".) >> >> The problem with fixing LabeledMolecule.name is that it is frozen. So >> the >> two options would be to either edit the ccp/molecule/LabeledMolecule xml >> file by hand or to do something like the following at the prompt: >> >>>>> labeledMolecule = >>>>> top.project.findFirstLabeledMolecule(name='p7:methanol_CN') >>>>> labeledMolecule.__dict__['name'] = 'p7' >> >> The GUI would not update (we are bypassing the normal API code) but it >> ought to save with that name and then reload ok. >> >> LabeledMolecule was in v1. >> >> >> If you go to http://www2.ccpn.ac.uk/download/ccpnmr/ you will find older >> releases. The data model version is indeed out of sync with Analysis >> versions. >> >> Wayne >> >> On Fri, 16 Nov 2012, Gary Thompson wrote: >> >>> Hi Rasmus >>> >>> Sorry more detail, I seem to be somethat confused how this came >>> about and >>> what to do... >>> >>> I seem to have a molecule and two labelled molecules: >>> >>> >>> print top.nmrProject.parent.labeledMolecules >>> >>> frozenset([<ccp.molecule.LabeledMolecule.LabeledMolecule >>> ['p7:methanol_CN']>,<ccp.molecule.LabeledMolecule.LabeledMolecule >>> ['p7:methanol_N']>]) >>> >>> print top.nmrProject.parent.molecules >>> >>> frozenset([<ccp.molecule.Molecule.Molecule ['p7']>]) >>> >>> The labelled molecules were (I believe) added as I set up samples in >>> the eci >>> interface (one is labelle dmehtanol_ N and the other methanol_CN) >>> surely they >>> should all refer to the same molecule (in actual fact they refer to >>> None) >>> >>> >>> print list(top.nmrProject.parent.labeledMolecules)[0].molecule >>> >>> None >>> >>> print list(top.nmrProject.parent.labeledMolecules)[1].molecule >>> >>> None >>> >>> >>> with names >>> >>> >>> print list(top.nmrProject.parent.labeledMolecules)[0].name >>> >>> p7:methanol_CN >>> >>> print list(top.nmrProject.parent.labeledMolecules)[1].name >>> >>> p7:methanol_N >>> >>> am I right in seeming to find that this is an eci problem? NB the >>> original >>> data model was imported from a version 1.0 model via the update >>> server, did >>> version 1.0 have a labelledMolecule? >>> >>> Finally how do a I correct the names of the labelled molecules? >>> surely they >>> now both have to be p7, but at the same time they are keyed by just >>> their >>> name and so you can's have two called p7... with different labelling... >>> >>> >>> Sorry, further questions: where do I find older versions of the >>> datamodel on >>> the ccpn web site? secondly the version of the model on the website >>> is 2.05, >>> I take it the data model version is out of sync with the analysis >>> software >>> versions? >>> >>> once again thanks for any help >>> >>> regards >>> gary >>> >>> -- >>> ------------------------------------------------------------------- >>> Dr Gary Thompson [Homans Lab Research Coordinator] >>> >>> Astbury Centre for Structural Molecular Biology, >>> University of Leeds, >>> Leeds, LS2 9JT, West-Yorkshire, UK Tel. +44-113-3433024 >>> email: [log in to unmask] Fax +44-113-3431935 >>> ------------------------------------------------------------------- >>> >> > -- Dr Andy Herbert Department of Chemistry University of Edinburgh West Mains Road Edinburgh UK EH9 3JJ Tel: +44 (0)131 650 4704 or 650 7372 Email: [log in to unmask] The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336.