MODelability Index (MODI) that estimates the feasibility of obtaining predictive QSAR models (correct classification rate above 0.7) for a binary data set of bioactive compounds. MODI is defined as an activity class-weighted ratio of the number of nearest-neighbor pairs of compounds with the same activity class versus the total number of pairs.
This program will run on all operating systems.
To run MODI, you need to:
- Clone this repository
- setup Python and third-party dependencies.
- (optionally) install this package.
See "Clone" button on this page for further information
Our package has been developed and tested using Python 3 and the following versions of the third-party packages:
- rdkit
- numpy
- pandas
- copy
- tqdm
Our package has been developed and tested using :
- MAACS
- Morgan (radius 2 and nBits 2048)
- Morgan4 (radius 4 and nBits 2048)
- Morgan_chiral2 (radius 2 and nBits 2048)
- Morgan_chiral2 (radius 4 and nBits 2048)
- SiRMS (Simplex (fragment) representation of molecular structure)
To see the help on how to run the program go to examples folder:
examples/Tutorial_classification.ipynb
For any suggestions, comments, issues, please contact us using "rodolpho-at-insilicall-dot-com"