Difference between revisions of "Tutorial for the relaxation dispersion auto-analysis in the GUI"
Jump to navigation
Jump to search
(→Tutorial: Formatting improvements.) |
|||
(14 intermediate revisions by 2 users not shown) | |||
Line 1: | Line 1: | ||
+ | __TOC__ | ||
+ | |||
== Tutorial == | == Tutorial == | ||
This follows the setup for test data in the Manual | This follows the setup for test data in the Manual | ||
Line 5: | Line 7: | ||
Where the test data is in: | Where the test data is in: | ||
− | + | test_suite/shared_data/dispersion/Hansen | |
It is written up here as a script instead. This goes a little faster testing. | It is written up here as a script instead. This goes a little faster testing. | ||
Line 16: | Line 18: | ||
</source> | </source> | ||
+ | Then we build the script onwards.<br> | ||
+ | |||
+ | We run relax repeatedly, to execute code. Then we write new code in the script, and run again. | ||
+ | <source lang="bash"> | ||
+ | relax test.py | ||
+ | </source> | ||
+ | When we are satisfied, one can then do like this. | ||
+ | |||
+ | Start relax | ||
+ | <source lang="bash"> | ||
+ | relax -g -t log.txt | ||
+ | </source> | ||
+ | |||
+ | Then do | ||
+ | User functions -> Script -> test.py | ||
+ | |||
+ | THEN do: | ||
+ | View -> Data pipe editor -> Right click on pipe -> Associate with a new Auto analysis | ||
+ | |||
+ | '''This should bring you to a window, where all settings have been set.''' | ||
+ | |||
+ | Set | ||
+ | * Relaxations dispersion models: ['R2eff', 'No Rex', 'CR72', 'NS CPMG 2-site expanded'] | ||
+ | * Grid increements: 11 (For speed-up in test phase) | ||
+ | * Monte-Carlo simulations number: 5 (For speed up in test phase) | ||
+ | |||
+ | Then a quick click on spin.isotope function, and GO. | ||
+ | |||
+ | === The script === | ||
+ | |||
+ | {{collapsible script | ||
+ | | type = Python script | ||
+ | | title = The test.py script. | ||
+ | | lang = python | ||
+ | | script = | ||
+ | # Python modules. | ||
+ | import os | ||
+ | import glob | ||
+ | from time import asctime, localtime | ||
+ | |||
+ | # relax modules. | ||
+ | from lib.io import sort_filenames | ||
+ | |||
+ | # Set path to data. | ||
+ | data = '/sbinlab2/tlinnet/software/NMR-relax/relax_trunk/test_suite/shared_data/dispersion/Hansen' | ||
+ | |||
+ | # Create the data pipe. | ||
+ | pipe_bundle = "relax_disp (%s)" % asctime(localtime()) | ||
+ | pipe_name = "origin - %s" % pipe_bundle | ||
+ | pipe.create(pipe_name=pipe_name, bundle=pipe_bundle, pipe_type='relax_disp') | ||
+ | |||
+ | # Create spin to hold data. | ||
+ | sequence.read(file='fake_sequence.in', dir=data, res_num_col=1, res_name_col=2) | ||
+ | deselect.read(file='unresolved', dir=data+os.sep+'500_MHz', spin_id_col=None, mol_name_col=None, res_num_col=1, boolean='AND', change_all=False) | ||
+ | deselect.read(file='unresolved', dir=data+os.sep+'800_MHz', spin_id_col=None, mol_name_col=None, res_num_col=1, res_name_col=2, boolean='AND', change_all=False) | ||
+ | |||
+ | # Give the spins attributes. | ||
+ | spin.isotope(isotope='15N', spin_id='@*', force=True) | ||
+ | spin.name(name='N') | ||
+ | |||
+ | # Do for 500. | ||
+ | ############# | ||
+ | |||
+ | # Change directory. | ||
+ | os.chdir(data + os.sep + '500_MHz') | ||
+ | |||
+ | # Get the file list, and sort the file list alphanumeric. | ||
+ | flist500 = glob.glob('*.in_sparky') | ||
+ | flist500 = sort_filenames(filenames=flist500) | ||
+ | |||
+ | # Make ID. | ||
+ | ID500 = [] | ||
+ | for f in flist500: | ||
+ | ID500.append("500_"+f.split(".in_sparky")[0]) | ||
+ | |||
+ | # Then read. | ||
+ | spectrum.read_intensities(file=flist500, spectrum_id=ID500) | ||
+ | |||
+ | # Repeat for the replicated spectra. | ||
+ | flist500rep = glob.glob('*in.bis_sparky') | ||
+ | flist500rep = sort_filenames(filenames=flist500rep) | ||
+ | |||
+ | # Make ID. | ||
+ | ID500rep = [] | ||
+ | for f in flist500rep: | ||
+ | ID500rep.append("500_"+f.split(".in.bis_sparky")[0]+'b') | ||
+ | |||
+ | # Then read. | ||
+ | spectrum.read_intensities(file=flist500rep, spectrum_id=ID500rep) | ||
+ | |||
+ | # Then map replicated. | ||
+ | for b_id in ID500rep: | ||
+ | a_id = b_id[:-1] | ||
+ | spectrum.replicated(spectrum_ids=[a_id, b_id]) | ||
+ | |||
+ | # Then check. | ||
+ | print(cdp.replicates) | ||
+ | |||
+ | # Then repeat for 800. | ||
+ | ###################### | ||
+ | |||
+ | # Change directory. | ||
+ | os.chdir(data + os.sep + '800_MHz') | ||
+ | |||
+ | # Get the file list, and sort the file list alphanumeric. | ||
+ | flist800 = glob.glob('*.in_sparky') | ||
+ | flist800 = sort_filenames(filenames=flist800) | ||
+ | |||
+ | # Make ID. | ||
+ | ID800 = [] | ||
+ | for f in flist800: | ||
+ | ID800.append("800_"+f.split(".in_sparky")[0]) | ||
+ | |||
+ | # Then read. | ||
+ | spectrum.read_intensities(file=flist800, spectrum_id=ID800) | ||
+ | |||
+ | # Repeat for the replicated spectra. | ||
+ | flist800rep = glob.glob('*in.bis_sparky') | ||
+ | flist800rep = sort_filenames(filenames=flist800rep) | ||
+ | |||
+ | # Make ID. | ||
+ | ID800rep = [] | ||
+ | for f in flist800rep: | ||
+ | ID800rep.append("800_"+f.split(".in.bis_sparky")[0]+'b') | ||
+ | |||
+ | # Then read. | ||
+ | spectrum.read_intensities(file=flist800rep, spectrum_id=ID800rep) | ||
+ | |||
+ | # Then map replicated. | ||
+ | for b_id in ID800rep: | ||
+ | a_id = b_id[:-1] | ||
+ | spectrum.replicated(spectrum_ids=[a_id, b_id]) | ||
+ | |||
+ | # Then check. | ||
+ | print(cdp.replicates) | ||
+ | print("%s %s %s %s" % (len(ID500), len(ID500rep), len(ID800), len(ID800rep))) | ||
+ | |||
+ | # Then set spectrum properties. | ||
+ | all_ID = ID500 + ID500rep + ID800 + ID800rep | ||
+ | |||
+ | for cur_id in all_ID: | ||
+ | # Split from name. | ||
+ | sfrq_str, vcpmg_str = cur_id.split("_") | ||
+ | |||
+ | if vcpmg_str == 'reference': | ||
+ | vcpmg = None | ||
+ | else: | ||
+ | vcpmg = float(vcpmg_str.split("b")[0]) | ||
+ | print("%s %s %s" % (cur_id, sfrq_str, vcpmg)) | ||
+ | |||
+ | # Set the current experiment type. | ||
+ | relax_disp.exp_type(spectrum_id=cur_id, exp_type='SQ CPMG') | ||
+ | |||
+ | # Set the NMR field strength of the spectrum. | ||
+ | spectrometer.frequency(id=cur_id, frq=float(sfrq_str), units='MHz') | ||
+ | |||
+ | # Relaxation dispersion CPMG constant time delay T (in s). | ||
+ | relax_disp.relax_time(spectrum_id=cur_id, time=0.03) | ||
+ | |||
+ | # Set the relaxation dispersion CPMG frequencies. | ||
+ | relax_disp.cpmg_setup(spectrum_id=cur_id, cpmg_frq=vcpmg) | ||
+ | }} | ||
+ | |||
+ | == Inspect results == | ||
+ | Go to the '''final''' folder. | ||
+ | |||
+ | Execute | ||
+ | <source lang="bash"> | ||
+ | cd $HOME/test/final | ||
+ | ./grace2images.py | ||
+ | </source> | ||
+ | Go through the PNG images | ||
+ | |||
+ | Also open | ||
+ | <source lang="bash"> | ||
+ | gedit $HOME/test/log.txt $HOME/test/final/chi2.out $HOME/test/final/models.out | ||
+ | gedit $HOME/test/No_Rex/chi2.out | ||
+ | gedit $HOME/test/CR72/chi2.out $HOME/test/CR72/kex.out | ||
+ | gedit $HOME/test/NS_CPMG_2-site_expanded/chi2.out $HOME/test/NS_CPMG_2-site_expanded/kex.out | ||
+ | </source> | ||
+ | |||
+ | == Get info from log.txt == | ||
+ | See [[Grep_log_file|Grep log file]] for inspiration. | ||
+ | |||
+ | Try these different grep commands | ||
+ | <source lang="bash"> | ||
+ | egrep -wi --color 'relax>| model -' $HOME/test/log.txt | ||
+ | </source> | ||
+ | |||
+ | Find '''eliminate''' function | ||
+ | <source lang="bash"> | ||
+ | egrep -wi --color -A 10 'relax> eliminate' $HOME/test/log.txt | ||
+ | </source> | ||
+ | |||
+ | Find '''model_selection''' function | ||
+ | <source lang="bash"> | ||
+ | egrep -wi --color -A 100 'relax> model_selection' $HOME/test/log.txt | ||
+ | </source> | ||
== See also == | == See also == | ||
[[Category:Tutorials]] | [[Category:Tutorials]] | ||
[[Category:Relaxation dispersion analysis]] | [[Category:Relaxation dispersion analysis]] |
Latest revision as of 18:37, 6 November 2015
Tutorial
This follows the setup for test data in the Manual
- The relaxation dispersion auto-analysis in the GUI
- http://www.nmr-relax.com/manual/The_relaxation_dispersion_auto_analysis_in_the_GUI.html
Where the test data is in:
test_suite/shared_data/dispersion/Hansen
It is written up here as a script instead. This goes a little faster testing.
In terminal
mkdir -p $HOME/test
cd $HOME/test
gedit test.py
Then we build the script onwards.
We run relax repeatedly, to execute code. Then we write new code in the script, and run again.
relax test.py
When we are satisfied, one can then do like this.
Start relax
relax -g -t log.txt
Then do
User functions -> Script -> test.py
THEN do:
View -> Data pipe editor -> Right click on pipe -> Associate with a new Auto analysis
This should bring you to a window, where all settings have been set.
Set
- Relaxations dispersion models: ['R2eff', 'No Rex', 'CR72', 'NS CPMG 2-site expanded']
- Grid increements: 11 (For speed-up in test phase)
- Monte-Carlo simulations number: 5 (For speed up in test phase)
Then a quick click on spin.isotope function, and GO.
The script
Python script: The test.py script.
Inspect results
Go to the final folder.
Execute
cd $HOME/test/final
./grace2images.py
Go through the PNG images
Also open
gedit $HOME/test/log.txt $HOME/test/final/chi2.out $HOME/test/final/models.out
gedit $HOME/test/No_Rex/chi2.out
gedit $HOME/test/CR72/chi2.out $HOME/test/CR72/kex.out
gedit $HOME/test/NS_CPMG_2-site_expanded/chi2.out $HOME/test/NS_CPMG_2-site_expanded/kex.out
Get info from log.txt
See Grep log file for inspiration.
Try these different grep commands
egrep -wi --color 'relax>| model -' $HOME/test/log.txt
Find eliminate function
egrep -wi --color -A 10 'relax> eliminate' $HOME/test/log.txt
Find model_selection function
egrep -wi --color -A 100 'relax> model_selection' $HOME/test/log.txt