DME Inhibitor Prediction via Descriptor Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for predicting the inhibiting character of drug metabolizing enzymes (DMEs) like CYP, SULT, and UGT are limited by the large number of descriptors used, which slows down computational time and fails to explain the inhibiting factors, and only consider a single binding energy conformation.
Innovation Solution
A method is developed to select a subset of molecular descriptors based on their relative importance, using physicochemical parameters and binding energies from multiple conformations, and trains a classification model like Random Forest or Support Vector Machine to predict inhibitor status, reducing computational time and improving understanding of inhibiting factors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a large number of molecular descriptors are used in the classification model, then the prediction accuracy may be improved, but the computational time increases significantly
Solution Approach 1:
The patent extracts and selects only the most relevant molecular descriptors from a large initial set based on their importance for predicting DME inhibition. This selective extraction maintains prediction accuracy while reducing the computational burden by eliminating redundant or less informative descriptors from the analysis.
Solution Approach 2:
The patent changes the parameter set by transforming a large number of molecular descriptors into a reduced subset of key descriptors. This parameter transformation maintains the essential information needed for accurate prediction while significantly reducing the computational complexity and processing time required.
2Measurement precision
If a large number of molecular descriptors are used in the classification model, then more comprehensive molecular characteristics are considered, but the model becomes more complex and harder to interpret
Solution Approach 1:
The patent extracts the most critical molecular descriptors that contribute significantly to DME inhibition prediction, removing redundant descriptors. This extraction process simplifies the model structure while preserving the essential molecular characteristics needed for accurate prediction, making the model both simpler and more interpretable.
Solution Approach 2:
The patent segments the large set of molecular descriptors into relevant and irrelevant groups, keeping only the essential ones for the final model. This segmentation approach breaks down the complex descriptor set into manageable, meaningful components that enhance model interpretability without sacrificing predictive power.
3Ease of operation
If only a single binding energy conformation is considered, then the computational process is simpler, but the prediction accuracy is reduced
Solution Approach 1:
The patent applies partial action by considering multiple binding energy conformations rather than just a single conformation. This partial extension beyond the minimal single-conformation approach captures the conformational flexibility of the enzyme-substrate interaction, improving prediction accuracy while maintaining computational feasibility through selective sampling of relevant conformations.
Data Source
AI summary
The present invention relates to the prediction of drug metabolizing enzymes (DME) inhibitors. Inhibition of DMEs leads to adverse drug-drug interaction, hence predicting inhibition of DMEs by determined molecules is critical for preventing drug toxicity. Inventors elaborated a protocol of integrated in silico protein structure-based and machine learning approach to predict inhibition of DMEs. In particular, the present invention relates to a method for training a model for predicting inhibition of DMEs, comprising a selection of a number of descriptors among an initial set comprising physicochemical descriptors and binding energies on at least one enzyme configuration, and a training of a classification model on a learning database of known inhibitors or non-inhibitors based on the selected descriptors as inputs. This approach successfully predicted inhibition of CYP2C9, CYP2D6, SULT1A1, SULT1A3 and UGT1 A1.