Wiki

Clone wiki

OOMMPPAA / LoadingData

Loading data

Example data is available in /data in the directory you install in.

You can load your own data into OOMMPPAA and perform the analysis from the UI in the windows version. Simply follow the instructions from the homepage. This process will load a new command prompt window that will run your request. Closing the window will kill the request. But don't worry - rerunning the same request will not duplicate anything.

We have supplied trial data to act as a template for future work. The files are found in the data directory. Running this with a new target name -> e.g. MYCDK2 or TRIAL or QWER will make a new target and load the data in accordingly. Give it a go!!!

Input files

mols.sdf is an SD File of all of the ligands for which you have structures. The structures need to be aligned into the same co-ordinate frame by aligning their complexed proteins. We recommend only using crystal structures, however in theory you could supplement this file with docked ligands.

acts.csv is a comma separated file of activity data. Fields are denoted using the header.

Required fields:

  • "smiles" is the isomeric smiles string for the molecule tested. Racemic data should not be entered.
  • "Activity" is the activity in -log(10) form. We publish work from OOMMPPAA using IC50 and Ki data only. We have also tested it using Tm Shift data, which showed some interesting results. However it is important to have confidence in the activity data you use.

Optional fields:

  • "Operator" is the operator on the activity data, less than indicates inactive compounds. If not entered it is assumed all values are "="
  • "ID" is the internal ID for the compound. This column can be left blank.

prot.pdb is a PDB file of your protein in the coordinate frame of the molecules. It should be apo with no water molecules or buffers. It should be in a conformation you would like to see when doing the analysis.

Processing

This processing can take some time for large data sets. If you have say 1,000 activity compounds and 30 crystal structures it may take an hour or more.

You can use the application while the data is loading.

Command line:

We offer a command line utility for loading your own data.

Go into the directory src/WebApp from where you have installed OOMMPPAA

cd ~/CHOC/src/WebApp
Then use manage.py to add data
python manage.py --acts $ACTS.csv --mols $MOLS.sdf --prot $PROT.pdb --mmp True

To get a full range of options type:

python manage.py --help

User interface:

  • Press the "Add Data" button ss.png
  • Add a new target
  • Type in the target name Untitled.png
  • Select the activity, structural (molecules) and protein files Untitled.png
  • Upload and process! Untitled.png
  • Now watch the processing happen - the warning message is normal! Untitled.png

Updated