Matrix Metalloproteinase 9 small molecule analysis and drug discovery
This set of jupyter notebooks and python scripts is used for a general analysis of biologically active small molecules against MMP9.
MMP9 is a proteinase in the extracellular matrix that is crucial for the correct modeling of extracellular matrix structures. As such MMP9 has been linked to various physiological diseases. Inhibition of the MMP9 function has been shown to decrease cancer progression among other extracellular matrix related and cell signaling induced diseases.
The ChEMBL database offers biologically relevant data on a wide range of drug-like small molecules interacting with various protein targets. This data analysis specifically aims to uncover the underlying properties of small molecule inhibitors for MMP9. This is done through an in depth data analysis of the small molecules and their chemical properties. The small molecules are then docked into the active site of MMP9 through an in vitro docking screen to give binding energy values. This leads to an eventual generation of machine learning models to establish a rule set for high affinity small molecule drugs against MMP9. The main goal is to provide an accurate bioinformatic examination and aid in the drug discovery for diseases related to this extracellular matrix proteinase.
The current model accuracy through the HistGradientBoostingRegressor machine learning model is sitting at 67.4% with an adjusted accuracy of 50.0% and an RMSE of 0.96.
Note: I am actively working on this project. If you have any comments or suggestions, do not hesitate to contact me. This includes inquiries on the written code and any suggestions for potential protein-ligand analyses.
Required Python packages:
numpy pandas chembl-webresource-client rdkit scipy seaborn matplotlib scikit-learn lazypredict