Abstract Molecular docking is an important step of the drug discovery process which aims at calculating the preferred position and shape of one molecule to a second when they are bound to each other. During such analysis, 3D representations of molecules are manipulated according to their degree of freedoms: rigid roto-translation and fragment rotations along the rotatable bonds. In our work, we focussed on one specific phase of the molecular docking procedure i.e. molecular unfolding (MU), which is used to remove the initial bias of a molecule by expanding it to an unfolded shape simpler to manipulate within the target cavity. The objective of the MU problem is to find the configuration that maximizes the molecular area, or equivalently, that maximizes the internal distances between atoms inside the molecule. We propose a quantum annealing approach to MU by formulating it as a high-order unconstrained binary optimization which was possible to solve on the latest D-wave annealing hardware (2000Q and advantage). Results and performances obtained with quantum annealers are compared with state of art classical solvers.
Abstract Molecular docking is an important step of the drug discovery process which aims at calculating the preferred position and shape of one molecule to a second when they are bound to each other. During such analysis, 3D representations of molecules are manipulated according to their degree of freedoms: rigid roto-translation and fragment rotations along the rotatable bonds. In our work, we focussed on one specific phase of the molecular docking procedure i.e. molecular unfolding (MU), which is used to remove the initial bias of a molecule by expanding it to an unfolded shape simpler to manipulate within the target cavity. The objective of the MU problem is to find the configuration that maximizes the molecular area, or equivalently, that maximizes the internal distances between atoms inside the molecule. We propose a quantum annealing approach to MU by formulating it as a high-order unconstrained binary optimization which was possible to solve on the latest D-wave annealing hardware (2000Q and advantage). Results and performances obtained with quantum annealers are compared with state of art classical solvers.
Molecular docking is a crucial phase in drug discovery, involving the precise determination of the optimal spatial arrangement between two molecules when they bind. The such analysis, the 3D structure of molecules is a fundamental consideration, involving the manipulation of molecular representations based on their degrees of freedom, including rigid roto-translation and fragment rotations along rotatable bonds, to determine the preferred spatial arrangement when molecules bind to each other. In this paper, quantum annealing based solution to solve Molecular unfolding (MU) problem, a specific phase within molecular docking, is explored and compared with a state-of-the-art classical algorithm named "GeoDock". Molecular unfolding focuses on expanding a molecule to an unfolded state to simplify manipulation within the target cavity and optimize its configuration, typically by maximizing molecular area or internal atom distances. Molecular unfolding problem aims to find the torsional configuration that increases the inter-atomic distance within a molecule, which also increases the molecular area. Quantum annealing approach first encodes the problem into a Higher-order Unconstrained Binary Optimization (HUBO) equation which is pruned to an arbitrary percentage to improve the time efficiency and to be able to solve the equation using any quantum annealer. The resultant HUBO is then converted to a Quadratic Unconstrained Binary Optimization equation (QUBO), which is easily embedded on a D-wave annealing Quantum processor.
Abstract Molecular docking is an essential step in the drug discovery process involving the detection of three-dimensional poses of a ligand inside the active site of the protein. In this paper, we address the Molecular Docking search phase by formulating the problem in quadratic unconstrained binary optimization terms, suitable for an annealing approach. We propose a problem formulation as a weighted subgraph isomorphism between the ligand graph and the grid of the target protein pocket. In particular, we applied a graph representation to the ligand embedding all the geometrical properties of the molecule including its flexibility, and we created a weighted spatial grid to the 3D space region inside the pocket. The proposed quantum annealing-based method for molecular docking achieves valid ligand placements. Compared to simulated annealing, quantum solvers sampled fewer but higher-quality solutions with lower root-mean-square deviation, demonstrating competitive performance within hardware limits.
With the continuous growth of quantum hardware and its increasing accessibility, there is a shift away from theoretical research towards real-world problems. One of the most promising applications of quantum computing is drug discovery, which is a complex and multi-disciplinary process. To increase the success probability of finding new drugs, we can introduce an in-silico step to virtual screen the drug candidates. In this paper, we address the molecular docking problem of virtual screening. In particular, we analyze the performance of a molecular docking application, which is modeled as a Quadratic Unconstrained Binary Optimization (QUBO) instance. We leveraged the computational power offered by different quantum computing models, gate-based and quantum annealing, to speed the solution of the QUBO formulation. Experimental results show how quantum annealer is still the best in class for our case study. Moreover, we report here the experience and the takeaways we obtained by executing quantum molecular docking on publicly accessible quantum hardware and software.
Molecular docking (MD) is a crucial task in drug design, which predicts the position, orientation, and conformation of the ligand when it is bound to a target protein. It can be interpreted as a combinatorial optimization problem, where quantum annealing (QA) has shown a promising advantage for solving combinatorial optimization. In this work, we propose a novel quantum molecular docking (QMD) approach based on a QA-inspired algorithm. We construct two binary encoding methods to efficiently discretize the degrees of freedom with an exponentially reduced number of bits and propose a smoothing filter to rescale the rugged objective function. We propose a new quantum-inspired algorithm, hopscotch simulated bifurcation (hSB), showing great advantages in optimizing over extremely rugged energy landscapes. This hSB can be applied to any formulation of an objective function under binary variables. An adaptive local continuous search is also introduced for further optimization of the discretized solution from hSB. Concerning the stability of docking, we propose a perturbation detection method to help rank the candidate poses. We demonstrate our approach on a typical data set. QMD has shown advantages over the search-based Autodock Vina and the deep-learning DIFFDOCK in both redocking and self-docking scenarios. These results indicate that quantum-inspired algorithms can be applied to solve practical problems in drug discovery even before quantum hardware become mature.
Molecular docking (MD) is a crucial task in drug design, which predicts the position, orientation, and conformation of the ligand when bound to a target protein. It can be interpreted as a combinatorial optimization problem, where quantum annealing (QA) has shown promising advantage for solving combinatorial optimization. In this work, we propose a novel quantum molecular docking (QMD) approach based on QA-inspired algorithm. We construct two binary encoding methods to efficiently discretize the degrees of freedom with exponentially reduced number of bits and propose a smoothing filter to rescale the rugged objective function. We propose a new quantum-inspired algorithm, hopscotch simulated bifurcation (hSB), showing great advantage in optimizing over extremely rugged energy landscapes. This hSB can be applied to any formulation of objective function under binary variables. An adaptive local continuous search is also introduced for further optimization of the discretized solution from hSB. Concerning the stability of docking, we propose a perturbation detection method to help ranking the candidate poses. We demonstrate our approach on a typical dataset. QMD has shown advantages over the search-based Autodock Vina and the deep-learning DIFFDOCK in both re-docking and self-docking scenarios. These results indicate that quantum-inspired algorithms can be applied to solve practical problems in the drug discovery even before quantum hardware become mature.
Molecular docking is an important step in drug discovery, enabling the evaluation of receptor-ligand affinity while reducing experimental costs and increasing the number of possible tests. However, the high computational cost associated with molecular docking remains a limiting factor that can restrict both the experimental precision and the scale of the problems being addressed. To improve the future applicability of molecular docking, recent works have proposed the use of quantum algorithms based on Gaussian Boson Sampling quantum computers and also gate-based quantum computers. In this work, we propose the use of Quantum Circuit Evolution (QCE) for solving the molecular docking problem, a gate based and gradient-free quantum evolutionary method whose evolution is driven by the random application of unitary operations to a quantum circuit. The proposed algorithm demonstrated the ability to find the best solution to the problem in fewer steps than the methods presented in previous studies, exhibiting fast and stable convergence.