The Challenge
Parkinson's Disease: Improving L-DOPA Therapy
Parkinson's disease is the second most common neurodegenerative disorder, affecting over 10 million people worldwide. The gold-standard treatment is L-DOPA, which is converted to dopamine in the brain. However, the enzyme catechol-O-methyltransferase (COMT) rapidly metabolizes L-DOPA to 3-O-methyldopa before it can reach the brain, significantly reducing its therapeutic efficacy. COMT inhibitors extend the half-life and bioavailability of L-DOPA, but currently marketed inhibitors such as entacapone, tolcapone, and opicapone suffer from toxicities and suboptimal pharmacokinetic profiles. There is a clear need for novel, safer COMT inhibitors with improved properties.
Our Approach
An integrated drug discovery pipeline from data assembly to in vivo validation
Data Assembly & Curation
Compiled a unified dataset of 395 unique chemical entities with COMT inhibitory activity, combining a proprietary NovaMechanics collection of 116 compounds with 279 compounds from ChEMBL. Compounds were classified as active (IC50 ≤ 1000 nM or ≥ 90% inhibition) or inactive based on their experimental activity profiles.
Ligand-Based Machine Learning
Built predictive ML models using Random Tree, kNN, and Random Forest algorithms. A consensus model achieved 82.4% accuracy with 84.9% sensitivity and 77.6% precision for predicting COMT inhibitory activity. These models were deployed on the Enalos Cloud Platform for accessible predictions.
Structure-Based Virtual Screening
Performed high-throughput virtual screening of 14,400 small molecules using molecular docking against the COMT crystal structure. Selected 30 commercially available compounds with satisfactory docking scores, followed by ROC analysis and polar score evaluation to prioritize candidates.
Molecular Dynamics & Binding Analysis
Ran 50ns molecular dynamics simulations to validate docking results and computed MM-PBSA and MM-GBSA binding free energies. Confirmed stable protein-ligand interactions and identified MHC01857 (Compound 12) as the most promising hit with the best binding energy profile (MM-GBSA: −31.3 kcal/mol).
In Vitro COMT Inhibition Testing
Screened 30 compounds at 25 μM against mouse COMT using an HPLC-based assay measuring 2-methoxyestradiol formation. Compound 12 was identified as the first hit, showing concentration-dependent inhibition (~53% at 50 μM). Lead optimization based on Compound 12 identified Compound 27, the most potent inhibitor at 29% inhibition at 10 μM.
IC50 Determination & In Vivo Testing
Determined IC50 values for Compound 27 on both mouse COMT (8.1 μM) and human COMT (19.9 μM), confirming cross-species activity. Toxicity testing via MTT assay showed no cytotoxicity up to 25 μM. In vivo testing in C57/Bl6 mice evaluated the compound's ability to inhibit peripheral L-DOPA metabolism to 3-O-methyldopa.
Key Results
Powered by NovaMechanics Software
Enalos Asclepios KNIME Nodes
Used for molecular dynamics simulations, structure-based virtual screening, and automated docking pipelines. Enabled MM-PBSA/MM-GBSA binding free energy calculations to validate and rank candidates.
Learn moreEnalos Cloud Platform & Enalos Suite
Hosted the COMT ligand-based predictive model as a web application for accessible screening. The Enalos Suite integrated COMT and ADMET models for comprehensive compound evaluation.
Learn moreRelated Publication
Cheminformatics and virtual screening studies of COMT inhibitors as potential Parkinson's disease therapeutics
This project identified a novel COMT inhibitor (Compound 27) with confirmed activity on both mouse and human COMT, representing a promising starting point for further optimization. While currently marketed COMT inhibitors operate in the nanomolar range, the identified compound demonstrates a new chemical scaffold with zero cytotoxicity and good tolerability in vivo. The project also delivered new in silico tools — including the Enalos COMT predictive cloud platform and ADMET models — that continue to serve the drug discovery community.
Have a Similar Challenge?
Let's explore how our computational drug discovery pipeline can accelerate your research.
Get in Touch