Skip to content

MDM2pred is a machine learning application based on the KNNRegressor algorithm, it's trained on 1647 known inhibitors of the human E3 ubiquitin ligase (Mouse Double Minute 2; MDM2), the primary negative regulator of the well-known tumor suppressor p53. The KNN model backing MDM2pred achieves ~0.74 R² on test compounds (cross-validated) and has a…

License

Notifications You must be signed in to change notification settings

naeemmrz/MDM2pred

Repository files navigation

MDM2pred

Logo

MDM2pred is a powerful machine learning tool for predicting the inhibitory potency of compounds against the human E3 ubiquitin ligase MDM2, a key regulator of the tumor suppressor p53. Based on the KNeighbors Regressor algorithm, MDM2pred has been trained on a comprehensive dataset of 1647 known MDM2 inhibitors, achieving an impressive R² value of ~0.74 and an RMSE of ~0.70 (in pIC50 units) over a 10-fold cross-validation. By simply inputting the SMILE notation of any compound, MDM2pred predicts its pIC50 value against MDM2 and returns the result as IC50. MDM2pred can be a valuable resource for researchers and drug developers looking to accelerate their early screening steps.

Authors

Naeem Abdul Ghafoor¹ & Ayşegül Yildiz¹² @Yildiz Neuro Lab

¹Department of Molecular Biology and Genetics, Graduate School of Natural and Applied Sciences, Mugla Sitki Kocman University, 48000 Mugla, Turkey
³Department of Molecular Biology and Genetics, Faculty of Science, Mugla Sitki Kocman University, Mugla, Turkey.

Usage

Open in Streamlit

  • To use MDM2pred, simply click on the "Open in Streamlit" badge above.
  • Enter the SMILES notation for your compound of interest.
  • Within seconds, MDM2pred will generate a 2D depiction of the compound and its predicted IC50 value against MDM2.
  • Use this information to gain valuable insights into the inhibitory potential of your compound and accelerate your drug development efforts.

To reproduce the results (Not available at the moment)

Download the project

  wget https://github.com/naeemmrz/mdm2pred.git

Unzip and enter the project directory

  unzip mdm2pred-main.zip
  cd mdm2pred-main

Install dependencies

  pip install -r requirements.txt

Run the Reproduce.py

  python Reproduce.py

The results will be printed out in the terminal.

Run Locally

Download the project

  wget https://github.com/naeemmrz/mdm2pred.git

Unzip and enter the project directory

  unzip mdm2pred-main.zip
  cd mdm2pred-main

Install dependencies

  pip install -r requirements.txt

Start the application

  streamlit run MDM2pred.py

The Application will open in your default browser with the same interface as the online version.

Acknowledgements

The development of MDM2pred was funded by the Research Support and Funding Office (BAP) of Mugla Sitki Kocman University under the project number 22/138/01/3/4.

About

MDM2pred is a machine learning application based on the KNNRegressor algorithm, it's trained on 1647 known inhibitors of the human E3 ubiquitin ligase (Mouse Double Minute 2; MDM2), the primary negative regulator of the well-known tumor suppressor p53. The KNN model backing MDM2pred achieves ~0.74 R² on test compounds (cross-validated) and has a…

Topics

Resources

License

Stars

Watchers

Forks

Languages