Computer implemented method and system for small molecule drug discovery

Abstract

In a small molecule drug discovery method, a transition state for a specific enzyme is modelled using quantum mechanics and molecular dynamics based simulation of the enzyme and substrate reaction; data defining the transition state (a ‘quantum pharmacophore’) is fed to a machine learning engine configured to generate transition state analogues, such as enzyme inhibitors.

1 . A small molecule drug discovery method comprising the steps of:
 (a) generating data representing a transition state for a specific enzyme, using quantum mechanics and molecular dynamics based simulation of the enzyme and substrate reaction, running on a computer-implemented simulation engine;   (b) storing the data defining the transition state (‘the quantum pharmacophore’) in a memory;   (c) passing the quantum pharmacophore into a computer-implemented machine learning engine configured to generate transition state analogues, in which transition state analogues are small molecules that enable the enzyme to enter its transition state.   
     
     
         2 . The method of  claim 1  in which the simulation engine generates data and information that allows for the development of simplified and computationally efficient descriptions of the transition state that form the quantum pharmacophore. 
     
     
         3 . The method of  claim 1  in which the quantum mechanics and molecular dynamics based simulation of the enzyme/substrate reaction uses a hybrid QM and classical model. 
     
     
         4 . The method of  claim 1  in which the quantum mechanics and molecular dynamics based simulation of the enzyme/substrate reaction combines a pure QM model and a QM/MM model. 
     
     
         5 . The method of  claim 1  in which the quantum mechanics and molecular dynamics based simulation of the enzyme/substrate reaction uses perturbative corrections. 
     
     
         6 . The method of  claim 1  in which the quantum mechanics and molecular dynamics based simulation of the enzyme/substrate reaction uses extensions of a molecular mechanics framework to reduce computational cost. 
     
     
         7 . The method of  claim 1  in which the quantum mechanics and molecular dynamics based simulation of the enzyme/substrate reaction uses extensions of a molecular mechanics framework using machine learning on QM simulation data. 
     
     
         8 . The method of  claim 1  in which the quantum mechanics based simulation of the enzyme/substrate reaction does not require ligand binding data, such as from wet lab screening experiments. 
     
     
         9 . The method of  claim 1  in which the quantum mechanics and molecular dynamics based simulation of the enzyme/substrate reaction only requires a crystal structure of the enzyme and knowledge of the reaction mechanism. 
     
     
         10 . The method of  claim 1  in which the quantum mechanics and molecular dynamics based simulation of the enzyme/substrate reaction only requires a crystal structure of the enzyme and knowledge of one or more of the following factors of the reaction mechanism: (i) input substrate(s) and co-factors, such as non-enzyme chemicals that impact the reaction, such as metal ions; (ii) output product(s) of the reaction; (iii) any allosteric binders required for the enzyme to be catalytically competent. 
     
     
         11 . The method of  claim 1  in which the quantum pharmacophore defines quantum mechanics data that is relevant to tight binding between the enzyme and substrate. 
     
     
         12 . The method of  claim 1  in which the quantum pharmacophore is a description of the structural and quantum chemical properties of the enzyme and substrate in the transition state. 
     
     
         13 . The method of  claim 1  in which the quantum pharmacophore captures the precise arrangement of atoms, groups, or functionalities in a small molecule required for specific interactions with its biological target and its activity. 
     
     
         14 . The method of  claim 1  in which the quantum pharmacophore is a template for the properties of a compound that would replicate the unique binding mode and high affinity of the substrate in the transition state to the enzyme. 
     
     
         15 . The method of  claim 1  in which the quantum pharmacophore is based on substrate features, such as transition state substrate features, constraints placed on the ligand by the enzyme configuration or a combination of both. 
     
     
         16 . The method of  claim 1  in which the machine learning engine is a generative ML system configured to optimise a cost function, namely similarity to the pharmacophore quantum mechanics transition state data. 
     
     
         17 . The method of  claim 1  in which the machine learning engine is a generative ML system configured to optimise a cost function that includes constraints based on medicinal chemistry input. 
     
     
         18 . The method of  claim 17  in which the medicinal chemistry input relates to the avoidance of specific chemical moieties, or divergence from the exact transition state pharmacophore by constraining properties such as polarity, charge distribution. 
     
     
         19 . The method of  claim 1  in which the machine learning engine uses a range of generative technologies, such as genetic algorithm based approaches to advanced graph based generative models. 
     
     
         20 . The method of  claim 19  in which the different generative technologies are selected depending on specific requirements. 
     
     
         21 . The method of  claim 19  in which the generative models are supported by predictive models for binding and PhysChem properties, enabling multi-parameter optimization towards desired chemical properties. 
     
     
         22 . The method of  claim 1  in which the machine learning engine uses the quantum pharmacophore to guide generative models to areas of chemical space that mimic the transition state and hence can inherit the properties of the transition state, including strong binding and a unique binding mode, leading to selectivity of compounds. 
     
     
         23 . The method of  claim 1  in which the machine learning engine uses feedback-driven generative models learn to recognize the quantum pharmacophore matching patterns and optimise interactions with the enzyme, as well as various druglike properties. 
     
     
         24 . The method of  claim 1  in which the machine learning engine identifies and discards transition state analogues which are not drug-like, have undesirable properties such as high toxicity or are too similar to the substrate or reactant. 
     
     
         25 . The method of  claim 1  in which outputs from the machine learning engine are assessed via wetlab assays and feedback loops to enable the optimization of candidates to provide designs of high potential inhibitors. 
     
     
         26 . The method of  claim 1  in which feedback loops include predictive models for other desirable properties. 
     
     
         27 . The method of  claim 1  in which intermediate wet lab results are fed back both into simulation and ML models to constantly improve the search and ultimately find the optimal lead series. 
     
     
         28 . The method of  claim 1  in which in silico and lab-based work is done on the machine learning engine outputs to filter out the transition state analogues which are not drug-like, have undesirable properties such as high toxicity or are too similar to the substrate or reactant. 
     
     
         29 . The method of  claim 1  in which in silico and lab-based work is done on the machine learning engine outputs to identify which transition state analogues have the tightest binding to the enzyme. 
     
     
         30 . The method of  claim 1  in which the best transition state analogues are refined to trade off binding strength with other properties which impact drug efficacy using another set of ML models. 
     
     
         31 . The method of  claim 1  in which the resultant compounds are developed into transition state drugs at the high-potential pre-clinical candidate stage. 
     
     
         32 . The method of  claim 1  in which the transition state analogues are de novo compounds. 
     
     
         33 . The method of  claim 1  in which the transition state analogues are identified using the virtual screening process. 
     
     
         34 . The method of  claim 1  in which the transition state analogue is an enzyme inhibitor. 
     
     
         35 . The method of  claim 1  in which the transition state analogue is a kinases inhibitor. 
     
     
         36 . The method of  claim 1  in which the transition state analogue is a protease. 
     
     
         37 . The method of  claim 1  in which the transition state analogue is a metallo protein. 
     
     
         38 . The method of  claim 1  in which the transition state analogue is a crop protection compound. 
     
     
         39 . A small molecule drug discovered by a method comprising the steps of:
 (a) generating data representing a transition state for a specific enzyme, using quantum mechanics and molecular dynamics based simulation of the enzyme and substrate reaction, running on a computer-implemented simulation engine;   (b) storing the data defining the transition state (‘the quantum pharmacophore’) in a memory;   (c) passing the quantum pharmacophore into a computer-implemented machine learning engine configured to generate transition state analogues, in which transition state analogues are small molecules that enable the enzyme to enter its transition state.   
     
     
         40 . (canceled) 
     
     
         41 . A small molecule drug virtual screening method comprising the steps of:
 (a) generating data representing a transition state for a specific enzyme, using quantum mechanics based simulation of the enzyme and substrate reaction, running on a computer-implemented simulation engine;   (b) storing the data defining the transition state (‘the quantum pharmacophore’) in a memory;   (c) passing the quantum pharmacophore into a computer-implemented machine learning engine configured to virtually screen known compounds to identify potential candidates as transition state analogues, in which transition state analogues are small molecules that enable the enzyme to enter its transition state.   
     
     
         42 . (canceled)