 |
 |
One more comment: I added some more planes to the dataset and it seems that the lack of peak selecting and assigning ability is specific to the last plane of the pseudo-3D. It also occurs for a case of 9 planes (in the ninth one). Patrick Patrick van der Wel wrote: > Hello, > > I wanted to post a follow up on this old message from last year, > regarding pseudo-3D data. I used the pipe2azara script to generate a > pseudo-3D azara file and read that into an existing analysis project. > That seemed to work fine and it recognized that it was a pseudo-3D and > even read out the parameter values that I had specified in the different > 2D slices of the nmrpipe files (using the sethdr program). However, > there were a couple of odd things I wanted to report. > > I was unable to assign any peaks in the third (and final) slice/plane of > the pseudo-3D. I was able to pick peaks in all three planes and was able > to use the 'a' shortcut to assign the first two planes, but there was no > response to me hitting 'a' in the third plane. Note that there were no > error messages on the command line. Note that I was also unable to just > select those picked peaks by just clicking on them (dragging around them > does work tho). > > The other thing I noticed is that it did not show up as an experiment > series in the rates analysis window. I had to manually click the 'edit > series' button, where I found out that not all information had been > entered for the experiment to be recognized as a proper series. > Specifically the condition type and units for the values. Perhaps it > would make sense to pop up that dialog box when a pseudo-3D is loaded > into the program? > > Otherwise things seem to work well and I managed to do some test rate > analysis on the data. By applying the group assign option I did manage > to get the third plane assigned, in a roundabout way. > > Patrick > > > > Wayne Boucher wrote: > >> Hello, >> >> Thanks to data and tips provided by Patrick van der Wel, I've now >> (hopefully correctly) added in support for pseudo 3D data (and in theory >> pseudo ND data) in NMRPipe. (Such support already existed for Azara >> format.) It seems that NMRPipe specifies the nucleus type to be ID for >> these pseudo dimensions. So now the program will detect this. >> >> The Analysis program still only reads the full NMRPipe single-file format, >> not the multi-file format. But to get at the values automatically you >> need to access the multi files (since only one value, for the first plane, >> is stored in the single file). So when you select the single file in >> OpenSpectrumPopup it will bring up a dialog asking if it really is pseudo >> 3D (and if not it will remove the offending dimension) and what the values >> of the planes are. You can either type in the values by hand, in which >> case you don't need access to the multi files at all, or let the values be >> fetched from the multi files after you specify the directory and template >> (with the %03d) for the files. >> >> (You will see that the "template" label in that dialog is a button. If >> you hit that button it will let you select a file in the relevant >> directory. Just select any of the multi files and hit ok. Then edit the >> name to replace the 001, or whatever, with %03d. This could try an be a >> bit cleverer and do that replacement automatically but I haven't put that >> in yet.) >> >> I have also updated the pipe2azara script, which is capable of dealing >> with the multi-file format, so that it automatically determines the values >> if you give it the multi-file template. This script will actually block >> the data, so has some advantage over importing the single-file NMRPipe >> data into Analysis. So, first convert the data to Azara format and then >> import into Analysis. (Well, for smallish data files this is no great >> advantage.) (I also fixed an unrelated bug in pipe2azara.) >> >> At some point we ought to add in a way of specifying that the other file >> formats, besides Azara and NMRPipe, are pseudo ND and bring up a similar >> dialog, except that the values would have to be definitely entered by >> hand. >> >> Wayne >> >>