CoChat
Sign Up
Cited by 232Quantum Computing Algorithms and ArchitectureDNA and Biological ComputingComputational Drug Discovery Methods

The prospects of quantum computing in computational molecular biology

Open access
Outeiral, Carlos, Strahm, Martin, Shi, Jiye, Morris, Garrett M., Benjamin, Simon C., Deane, Charlotte M.·2021·WIREs Computational Molecular Science
DOI 10.1002/wcms.1481
Open this paper in CoChat — free accountOpen PDF

Abstract

Quantum computers can in principle solve certain problems exponentially more quickly than their classical counterparts. We have not yet reached the advent of useful quantum computation, but when we do, it will affect nearly all scientific disciplines. In this review, we examine how current quantum algorithms could revolutionize computational biology and bioinformatics. There are potential benefits across the entire field, from the ability to process vast amounts of information and run machine learning algorithms far more efficiently, to algorithms for quantum simulation that are poised to improve computational calculations in drug discovery, to quantum algorithms for optimization that may advance fields from protein structure prediction to network analysis. However, these exciting prospects are susceptible to “hype,” and it is also important to recognize the caveats and challenges in this new technology. Our aim is to introduce the promise and limitations of emerging quantum computing technologies in the areas of computational molecular biology and bioinformatics. This article is categorized under: Structure and Mechanism > Computational Biochemistry and Biophysics Data Science > Computer Algorithms and Programming Electronic Structure Theory > Ab Initio Electronic Structure Methods

AI snapshot

Written from the paper’s full text.

TL;DR

This review surveys how quantum algorithms could affect computational molecular biology, especially statistical/machine learning, quantum simulation of electronic structure, and optimization. It argues that quantum simulation is likely the first useful application, while machine learning and optimization face serious input/output, hardware, and algorithmic caveats.

Key findings

  • Quantum machine learning could offer complexity advantages—e.g., HHL solves certain linear systems in \(O(\log N)\) versus \(O(N^3)\) classical matrix inversion—but extracting full solutions can require \(O(2^N)\) measurements, and most algorithms assume an unavailable QRAM.
  • A human genome (3 billion base pairs, \(1.2 \times 10^{10}\) classical bits, ~1.5 GB) could in principle be stored in ~34 qubits; 68 qubits would suffice for the genomes of every living human, though the state must be efficiently prepared.
  • Quantum simulation may solve the fully correlated electronic structure problem exactly within a basis set; a minimal-basis FCI calculation on the FeMo cofactor was estimated to require several months and circa 200 million high-quality qubits, though algorithmic advances lowered time requirements by orders of magnitude within 3 years.
  • Protein lattice folding is NP-hard, and the paper states quantum computers are not expected to give exponential speedup for NP-complete or harder problems; experimental work used D-Wave 2000Q for chignolin (10 residues) and Trp-Cage (8 residues), with evidence for only “limited quantum speedup.”

Methods

  • Narrative review, not a primary study; it introduces quantum information processing and then surveys three areas: statistical methods/machine learning, efficient simulation of quantum systems, and optimization problems.
  • Synthesizes published quantum algorithms, complexity comparisons, hardware error-rate reports, and protein-folding/optimization examples; no new dataset or experimental sample is analyzed.
  • The paper explicitly notes it leaves aside topics such as string algorithms for sequence analysis, medical imaging algorithms, numerical algorithms for differential equations, and biological network analysis.

Limitations

  • The authors state that space limitations forced them to omit several relevant areas, including sequence analysis, medical imaging, differential equations, and network analysis.
  • The authors emphasize practical caveats: input preparation and QRAM are unsolved, output extraction can erase speedups, appropriate data may be scarce, and fault-tolerant hardware requires large qubit overheads—e.g., surface-code error correction at 0.1% error could require several million qubits for 2,000-bit factorization versus ~4,000 noiseless qubits.
  • Evident from the text: as a non-systematic review, it does not define formal search or inclusion criteria, and many performance estimates are theoretical or prospective and may change with hardware and algorithmic advances.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1002/wcms.1481

Cites

Statistics

Nothing to recompute.

Claims the paper makes

0 of 114 cited sources checked · 0 supported · 0 flagged

  • Findingcitation

    The least error-prone gates to date have been reported in a trapped ion processor, with error rates of one part in 10^6 for single-qubit gates and errors of 0.1% for two-qubit gates.

    “The least error-prone gates to date have been reported in a trapped ion processor, with error rates of one part in 10 6 for single-qubit gates and errors of 0.1% for two-qubit gates [46,49].”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [46] source

      Not checked in the free run.

    • cites [49] source

      Not checked in the free run.

  • Findingcitation

    The current record for a controllable, programmable quantum processor is 53 qubits.

    “Given that the current record for a controllable, programmable quantum processor is 53 qubits [20], there is still a long way to go.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [20] source

      Not checked in the free run.

  • Premisecitation

    The existence of entanglement is fundamental for useful quantum computation.

    “The existence of entanglement is fundamental for useful quantum computation.”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [44]

      Not checked in the free run.

    • cites [45] source

      Not checked in the free run.

  • Premisecitation

    Technology to simulate quantum systems promises accurate predictions of drug-receptor interaction.

    “Finally, technology to simulate quantum systems has been developed which promises accurate predictions of drug-receptor interaction [18]”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [18] source

      Not checked in the free run.

  • Premisenone given

    Quantum computers would enable more effective algorithms by introducing operations that are not possible in classical machines.

    “A quantum computer would enable more effective algorithms by introducing operations that are not possible in classical machines.”
    Quote in the paperSubstantial claim
  • May combine two claims. 45
  • Classified as probably not a substantial claim. 38

  • May combine two claims.

    Findingcitation

    Google reported fidelities of 0.1% for single-qubit gates and 0.3% for two-qubit gates in a superconducting processor.

    “For comparison, the recent work on quantum supremacy by Google reported fidelities of 0.1% for single-qubit gates and 0.3% for two-qubit gates in a superconducting processor [20].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [20] source

      Not checked in the free run.

  • Classified as probably not a substantial claim.

    Premisereasoning

    The ability of a quantum computer to store and perform calculations on vast amounts of information-of the order of 2^N-by manipulating a few elements-of order N-forms the basis of its ability to provide a potentially exponential advantage over classical computers.

    “The ability of a quantum computer to store and perform calculations on vast amounts of information-of the order of 2 N -by manipulating a few elements-of order N-forms the basis of its ability to provide a potentially exponential advantage over classical computers.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationcitation

    Implementing most quantum gates will take exponentially large amounts of time, and may obliterate any quantum advantage.

    “Even if these gates can be used to solve useful problems, implementing them will take exponentially large amounts of time, and may obliterate any quantum advantage.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisenone given

    All quantum hardware proposals face a common challenge: errors during computation, which can ruin the computational process.

    “However, all of them face a common challenge: errors during computation, which can ruin the computational process.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Premisenone given

    Quantum processors do not work faster than classical computers, but operate in a fundamentally different way, achieving speedups by avoiding unnecessary computation.

    “Quantum processors do not work faster than classical computers, but operate in a fundamentally different way, achieving unprecedented speedups by avoiding unnecessary computation.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S4 (based).

    Premisenone given

    Quantum computers can in principle solve certain problems exponentially more quickly than their classical counterparts.

    “Quantum computers can in principle solve certain problems exponentially more quickly than their classical counterparts.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Premisereasoning

    The ability of quantum computers to apparently manipulate vast amounts of information with few operations is one of the cornerstones of quantum algorithms, and underpins their capacity to solve problems exponentially faster.

    “The ability of quantum computers to apparently manipulate vast amounts of information with few operations is one of the cornerstones of quantum algorithms, and underpins their capacity to solve problems exponentially faster.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Early processors leveraging quantum effects to perform classically impossible calculations are expected within the next decade.

    “Early processors leveraging quantum effects to perform classically impossible calculations are expected within the next decade [22].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [22] source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S4 (based).

    Premisenone given

    The best algorithms for protein folding, binding affinity, and large-scale genomic alignments require computational resources beyond the most powerful supercomputers.

    “The best algorithms for problems like predicting the folding of a protein, calculating the binding affinity of a ligand for a macromolecule, or finding optimal large-scale genomic alignments require computational resources that are beyond even the most powerful supercomputers of our era.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S4 (based).

    Premisenone given

    Quantum computers promise to solve biologically relevant problems, like predicting protein-ligand interactions or understanding coevolution of amino acids in proteins, exponentially faster than any current computer.

    “Quantum computers promise to solve biologically relevant problems, like predicting protein-ligand interactions or understanding coevolution of amino acids in proteins, exponentially faster than any current computer.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Premisecitation

    Rebentrost et al. presented a quantum algorithm that can train a SVM with a polynomial kernel in O(log N), later extended to the radial basis function kernel.

    “Rebentrost et al. [75] presented a quantum algorithm that can train a SVM with a polynomial kernel in O log N ð Þ, and this was later extended to the radial basis function (RBF) kernel [76].”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [75] source

      Not checked in the free run.

    • cites [76]

      Not checked in the free run.

  • May combine two claims.

    Interpretationreasoning

    Since it takes O(log N) to train a model, but O(N) calls to reconstruct the matrix, for a sufficiently large dataset it is impossible to recover a significant portion of the information, while it is still possible to train a model efficiently.

    “Since it takes O log N ð Þto train a model, but O N ð Þ calls to reconstruct the matrix, for a sufficiently large dataset it is impossible to recover a significant portion of the information, while it is still possible to train a model efficiently.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

    Premisenone given

    The input of information to a quantum computer is a much more serious problem.

    “The input of information to a quantum computer is a much more serious problem.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Premisenone given

    Quantum error correcting codes require using a large number of real, physical qubits to encode an abstract logical qubit, and this overhead becomes worse as the error rate increases.

    “Unfortunately, quantum error correcting codes require using a large number of real, physical qubits to encode an abstract logical qubit that is used for computation, and this overhead becomes worse as the error rate increases.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

    Premisenone given

    Most quantum machine learning algorithms assume the quantum computer has access to the dataset as a superposition state.

    “Most quantum machine learning algorithms assume the quantum computer has access to the dataset as a superposition state”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Premisecitation

    Implementing a quantum circuit to prepare a random quantum state can be very challenging, since approximating a random quantum state would be exponentially hard and wipe out any conceivable quantum advantage.

    “However, implementing it can be very challenging, since we expect that approximating a random quantum state would be exponentially hard and wipe out any conceivable quantum advantage [39].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [39] source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Gaussian process regression is commonly used to build surrogate models, for example in Bayesian optimization.

    “Gaussian process (GP) regression is a method commonly used to build surrogate models, for example in Bayesian optimization [125].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [125]

      Not checked in the free run.

  • May combine two claims.

    Premisereasoning

    Unlike the acceleration provided by GPUs, which speed up computation through massive parallelism, quantum algorithms have a complexity advantage: fewer operations are required to achieve the same task.

    “The key insight is that, unlike the acceleration provided by GPUs, which speed up computation through massive parallelism, quantum algorithms have a complexity advantage: fewer operations are required to achieve the same task.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Premisenone given

    Quantum error correcting codes require a very large increase in the number of qubits, so significant engineering advances are necessary before fault-tolerance can be achieved.

    “Unfortunately, these codes also require a very large increase in the number of qubits, so significant engineering advances are necessary before fault-tolerance can be achieved.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Premisenone given

    Quantum computers can in principle be far faster and able to handle far more data than their classical counterparts for machine learning.

    “However, they can in principle be far faster and able to handle far more data than their classical counterparts.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    A quantum algorithm for performing PCA was proposed by Lloyd et al.

    “A quantum algorithm for performing PCA was proposed by Lloyd et al. [86].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [86] source

      Not checked in the free run.

  • May combine two claims.

    Premisereasoning

    A single gate failure could corrupt a sensitive calculation, and error propagation can render the computation meaningless after only a small sequence of gates.

    “Given that a single gate failure could corrupt a sensitive calculation, it is not difficult to see that error propagation can render the computation meaningless after only a small sequence of gates.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Premisereasoning

    Although on paper quantum machine learning algorithms can vastly outperform classical analogues, practical difficulties still remain.

    “Unfortunately, although on paper quantum machine learning algorithms can vastly outperform classical analogues, practical difficulties still remain.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Premisecitation

    The quantum algorithm for prime number factorization could, in noiseless conditions, factorize a 2,000-bit number employing approximately 4,000 qubits and, assuming a 16 GHz gate rate, about a day to run.

    “For example, the quantum algorithm for prime number factorization [55] could, in noiseless conditions, factorize a 2,000-bit number employing approximately 4,000 qubits and, assuming a 16 GHz gate rate, about a day to run.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [55] source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    SVMs have been used extensively in the pharmaceutical industry for classification of small molecule data.

    “SVMs have been used extensively in the pharmaceutical industry for classification of small molecule data [124].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [124]

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    Although major applications will require very large, fault-tolerant quantum computers, near-term devices available within the next decade will be able to solve useful problems.

    “Although major applications will require very large, fault-tolerant quantum computers, it is expected that near-term devices available within the next decade will be able to solve useful problems [22].”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [22] source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    The quantum PCA algorithm builds the covariance matrix of the data in the quantum computer, and uses quantum phase estimation to compute the eigenvectors in logarithmic time.

    “In essence, this algorithm builds the covariance matrix of the data in the quantum computer, and uses a subroutine known as quantum phase estimation [15] to compute the eigenvectors in logarithmic time.”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [15]

      Not checked in the free run.

    • cites [86] source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    GPs are widely used to create QSAR models of drug properties and recently also for molecular dynamics simulations.

    “GPs are also widely used to create quantitative structure-activity relationship (QSAR) models of drug properties [126], and recently also for molecular dynamics simulations [127].”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [126]

      Not checked in the free run.

    • cites [127]

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Supervised learning has been used widely in biology for diverse problems including predicting the binding affinity of a ligand to a protein and computer-aided disease diagnosis.

    “Supervised learning has been used widely in biology, for problems as diverse as predicting the binding affinity of a ligand to a protein [122] and computer-aided disease diagnosis [123].”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [122]

      Not checked in the free run.

    • cites [123]

      Not checked in the free run.

  • May combine two claims.

    Premisenone given

    The best classical algorithms for persistent homology have a cost that is exponential in the dimensionality of the problem, which has limited its application.

    “Unfortunately, the best classical algorithms have a cost that is exponential in the dimensionality of the problem, which has limited its application.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Premisecitation

    Although some blueprints for QRAM have been proposed, there is not, to our knowledge, yet a working device.

    “Although some blueprints have been proposed [111,112], there is not, to our knowledge, yet a working device.”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites [111] source

      Not checked in the free run.

    • cites [112]

      Not checked in the free run.

  • May combine two claims.

    Interpretationnone given

    In principle, a quantum computer is able to efficiently solve the fully correlated electronic structure problem (the FCI equations), which would be a first step to accurately estimate binding energies and simulate the dynamics of chemical systems.

    “In principle, a quantum computer is able to efficiently solve the fully correlated electronic structure problem (the FCI equations), which would be a first step to accurately estimate binding energies, but also to simulate the dynamics of chemical systems.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    The quantum Boltzmann machine is expected to be more expressive.

    “Further work has presented a generalization, the quantum Boltzmann machine [134,135], which is expected to be more expressive.”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [134]

      Not checked in the free run.

    • cites [135]

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    Qubitization and quantum signal processing have provably optimal asymptotic scaling, although it is unclear whether this will translate to better practical applications.

    “There are many other Hamiltonian simulation methods, an example of a powerful and sophisticated method being qubitization [166] and quantum signal processing [167], which have the provably optimal asymptotic scaling, although it is unclear whether this will translate to better practical applications.”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [166] source

      Not checked in the free run.

    • cites [167]

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Interpretationnone given

    The major challenge in designing a quantum analogue of a neural network is how to implement a nonlinear activation function employing linear quantum gates.

    “The major challenge in designing a quantum analogue is how to implement a nonlinear activation function employing linear quantum gates.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Interpretationnone given

    The scaling advantage or disadvantage of training a multilayer perceptron in a quantum computer is unclear.

    “The scaling advantage or disadvantage of training a multilayer perceptron in a quantum computer is unclear, although the focus has been in the possible increased expressivity of these models.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

    Interpretationnone given

    These approaches aim to implement a quantum neural network expected to be more expressive than a classical neural network due to the greater power of quantum information.

    “These approaches aim to implement a quantum neural network, which is expected to be more expressive than a classical neural network due to the greater power of quantum information.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Multiple layers of artificial neural networks can discover intricate structures in raw data.

    “Recent developments in machine learning have been catalyzed by the discovery that multiple layers of artificial neural networks can discover intricate structures in raw data [108].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [108] source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    Multiple recent proposals for quantum neural networks involve measurements, dissipative quantum gates, continuous-variable quantum computing, and additional qubits to build linear gates that emulate nonlinearity.

    “There have been multiple recent proposals, and some ideas involve measurements [91][92][93], dissipative quantum gates [92], continuous-variable quantum computing [95], and the introduction of additional qubits to build linear gates that emulate nonlinearity [94].”
    Quote in the paperNot a substantial claim0 of 3 sources checked
    • cites [91][92][93]

      Not checked in the free run.

    • cites [92]

      Not checked in the free run.

    • cites [95] source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    HMM is one of the most common methods in computational biology, used widely in computational gene annotation and sequence alignment.

    “One of the most common methods in computational biology is the HMM, used widely in computational gene annotation and sequence alignment [59].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [59]

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S2 (checkable). May combine two claims.

    Interpretationreasoning

    An accurate and efficient algorithm to study electronic structure would enable better understanding of biological processes and open the door to next-generation engineering of biological interactions.

    “An accurate and efficient algorithm to study electronic structure would enable better understanding of biological processes, as well as open the door to next-generation engineering of biological interactions.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationnone given

    Quantum processors can potentially solve the electronic problem by direct diagonalization of the FCI matrix, yielding the exact result within a particular basis set, and thus solving problems arising from an incorrect description of the physics of molecular processes.

    “By contrast, quantum processors can potentially solve the electronic problem by direct diagonalization of the FCI matrix, yielding the exact result within a particular basis set, and thus solving a myriad of problems arising from an incorrect description of the physics of molecular processes (e.g.ligand polarization).”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    VQE is based on the Ritz-Rayleigh variational principle, which states that the expected energy of an approximate wavefunction will always be higher than the true ground state energy.

    “This algorithm, which underpins a great many other hybrid algorithms, is based on the Ritz-Rayleigh variational principle of quantum mechanics [14], which states that the expected energy of an approximate wavefunction will always be higher than the true ground state energy.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [14]

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    A heuristic to efficiently train quantum Boltzmann machines has been proposed that relies on system thermalization.

    “Recently, a heuristic to efficiently train quantum Boltzmann machines has been proposed that relies on system thermalization [105].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [105]

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    Deep learning successes in computational biology include accurate contact prediction in proteins, improved diagnostics for several diseases, molecular design and simulation.

    “Deep learning has begun to permeate every scientific discipline, and in computational biology its successes include accurate contact prediction in proteins [61], improved diagnostics for several diseases [128], molecular design [129] and simulation [130,131].”
    Quote in the paperNot a substantial claim0 of 3 sources checked
    • cites [61] source

      Not checked in the free run.

    • cites [128] source

      Not checked in the free run.

    • cites [129] source

      Not checked in the free run.

  • May combine two claims.

    Interpretationnone given

    It is very difficult to extract information from a quantum computer; retrieving the entire wavefunction is harder than computing the solution classically in the first place.

    “As discussed above, it is very difficult to extract information from a quantum computer.Although obtaining the energy is in general simple, retrieving the entire wavefunction is harder than computing the solution classically in the first place.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationreasoning

    Unlike classical approaches, quantum processors do not necessarily require an iterative process, although the initial guess still plays an important role.

    “Moreover, unlike classical approaches, they do not necessarily require an iterative process, although as we will discuss the initial guess still plays an important role.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Premisecitation

    Zhao et al. proposed use of the HHL algorithm for linear systems to invert the covariance matrix, achieving an exponential speedup as long as the matrix is sparse and well-conditioned, properties often achieved by covariance matrices.

    “Zhao et al. [69] proposed the use of the HHL algorithm for linear systems to invert this matrix, achieving an exponential speedup as long as the matrix is sparse and well-conditioned, properties that are often achieved by covariance matrices.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [69]

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based). May combine two claims.

    Interpretationnone given

    The quantum computers that can simulate large chemical systems will likely be fault-tolerant and therefore able to run arbitrarily deep algorithms without error.

    “The quantum computers that can simulate large chemical systems will likely be fault-tolerant and therefore able to run arbitrarily deep algorithms without error.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisenone given

    Grid-based simulations have the advantage of not requiring the Born-Oppenheimer approximation or a basis set.

    “Grid-based simulations have the advantage of not requiring the Born-Oppenheimer approximation or a basis set.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Premisecitation

    Full configuration interaction scales exponentially with the number of electrons, and even molecules with a handful of carbon atoms are barely accessible to computational investigation.

    “The most efficient algorithm to compute the energy of a system of electrons, full configuration interaction (FCI), scales exponentially with the number of electrons [14], and even molecules with a handful of carbon atoms are barely accessible to computational investigation [136].”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites [14]

      Not checked in the free run.

    • cites [136]

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    Grid-based methods are not naturally antisymmetric, which is required by the Pauli principle, so it is necessary to ensure antisymmetry using a sorting procedure.

    “However, they are not naturally antisymmetric, which is required by the Pauli principle of quantum mechanics, so it is necessary to ensure antisymmetry using a sorting procedure [143,161].”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [143] source

      Not checked in the free run.

    • cites [161]

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Some papers have proposed quantum implementations of generative adversarial networks.

    “Moreover, some papers have proposed quantum implementations of generative adversarial networks, another very popular model [98][99][100].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [98][99][100]

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S2 (checkable).

    Interpretationreasoning

    Quantum HMMs may lead to richer applications across computational biology.

    “This may lead to richer applications across computational biology.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationcitation

    In comparison with machine learning, the advantages of quantum simulation vastly outweigh the shortcomings, and quantum simulation is expected to be one of the first useful applications of practical quantum computing.

    “In comparison with machine learning, however, the advantages vastly outweigh the shortcomings, and it is expected that quantum simulation will be one of the first useful applications of practical quantum computing [22].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [22] source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    Finding the lowest-energy conformation of a lattice model is an NP-hard problem, meaning that under standard hypotheses no polynomial-time classical algorithm for this problem exists.

    “Finding the lowest-energy conformation of a lattice model is an NP-hard problem [205,206], meaning that under standard hypotheses, no polynomial-time classical algorithm for this problem exists.”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites [205]

      Not checked in the free run.

    • cites [206]

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    Random ansätze tend to lead to vanishing gradients when the number of qubits increases, meaning that it will be impossible to find a physically meaningful energy minimum.

    “While hardware-efficient ansätze are attractive due to their ease of implementation in current devices, recent work by McClean et al. has provided analytical and numerical evidence that random ansätze tend to lead to vanishing gradients when the number of qubits increases [174], meaning that it will be impossible to find a physically meaningful energy minimum.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [174] source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S4 (based).

    Interpretationnone given

    Early quantum processors that can solve useful problems are expected to arise within the next decade.

    “Early quantum processors that can solve useful problems are expected to arise within the next decade.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    A recent study by Outeiral et al. concluded that there is evidence for a limited quantum speedup, although this result may require adiabatic machines using error correction or quantum simulation in fault-tolerant universal machines.

    “A recent study by Outeiral et al. has employed numerical simulations to investigate this fact, concluding that there is evidence for a limited quantum speedup, although this result may require adiabatic machines using error correction or quantum simulation in fault-tolerant universal machines [209].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [209]

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    It is believed that adiabatic quantum computing, and quantum computing in general, is unable to efficiently solve the class of NP-complete problems, or at least no such feat has withstood scrutiny.

    “It is believed that adiabatic quantum computing (and quantum computing, in general) is unable to efficiently solve the class of NP-complete problems, or at least no such feat has withstood scrutiny [17,186].”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [17] source

      Not checked in the free run.

    • cites [186] source

      Not checked in the free run.

  • May combine two claims.

    Premisereasoning

    D-Wave Systems employs stoquastic Hamiltonians and is therefore nonuniversal.

    “The most popular experimental implementation of adiabatic quantum computing, commercialized by the company D-Wave Systems Inc., employs stoquastic Hamiltonians [180], and it is therefore nonuniversal.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [180]

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    Despite concerns, D-Wave has been used widely as an optimization metaheuristic and has recently been shown to overperform simulated annealing.

    “Despite these concerns, it has been used widely as an optimization metaheuristic and has recently been shown to overperform simulated annealing [188].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [188]

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    QAOA is a variational algorithm for optimization in a near-term quantum computer that has gathered considerable interest in the literature.

    “The quantum approximate optimization algorithm (QAOA) [189][190][191] is a variational algorithm for optimization in a near-term quantum computer that has gathered considerable interest in the literature.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [189][190][191]

      Not checked in the free run.

  • Classified as probably not a substantial claim. May combine two claims.

    Premisecitation

    It is currently believed that quantum computers cannot offer an exponential speedup to NP-complete and harder problems, although they can offer scaling advantages known as limited quantum speedup.

    “Furthermore, it is currently believed that quantum computers cannot offer an exponential speedup to NP-complete and harder problems [207], although they can offer scaling advantages that have been known in the literature as “limited quantum speedup” [208].”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites [207]

      Not checked in the free run.

    • cites [208] source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    There are concerns that stoquastic quantum computing may be classically simulatable, meaning that an exponential speedup may be impossible.

    “There are some concerns in the literature that this variety of quantum computing may be classically simulatable [17], meaning that an exponential speedup may be impossible.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [17] source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    According to the funnel hypothesis of protein folding, the native structure of a protein is believed to be the global minimum of its free energy, although many counterexamples exist.

    “According to the funnel hypothesis of protein folding, the native structure of a protein is believed to the global minimum of its free energy [177,193], although many counterexamples exist.”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [177] source

      Not checked in the free run.

    • cites [193]

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Reference [210] attempted to use quantum annealing to perform rotamer sampling in the Rosetta energy function [211].

    “[210] has attempted to use quantum annealing to perform rotamer sampling in the Rosetta energy function [211].”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [210]

      Not checked in the free run.

    • cites [211]

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    A very similar approach was presented by Marchand et al. [212] for conformer sampling.

    “A very similar approach was presented by Marchand et al. [212] for conformer sampling.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [212]

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S4 (based). May combine two claims.

    Premisenone given

    A quantum computer can store and manipulate states which, if represented classically, would correspond to vast amounts of information, and execute some algorithms that are exponentially faster than any classical alternative.

    “A quantum computer can store and manipulate states which, if represented classically, would correspond to vast amounts of information, and execute some algorithms that are exponentially faster than any classical alternative.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    Adiabatic quantum computing is based on the adiabatic theorem, which states that if a system is prepared in the ground state of a Hamiltonian and the Hamiltonian is varied slowly enough, the system will always remain in its instantaneous ground state.

    “Adiabatic quantum computing [181] is based on the adiabatic theorem of quantum mechanics [182].According to this theorem, if a system is prepared in the ground state of a Hamiltonian, and this Hamiltonian is varied slowly enough, the system will always remain in its instantaneous ground state.”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [181]

      Not checked in the free run.

    • cites [182] source

      Not checked in the free run.

70 of 70

  • The authors state this as an assumption.

    The quantum state encoding a dataset can be efficiently prepared.

    “as long as the state can be efficiently prepared (a considerable assumption, as later discussed)”

  • The authors state this as an assumption.

    A molecular Hamiltonian can be split as a sum of terms describing one- and two-electron interactions.

    “Trotterization is based on the premise that a molecular Hamiltonian can be split as a sum of terms describing one-and two-electron interactions.”
2 of 2

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

  • Quantum entanglement is essential to perform quantum computing.

    “It is well established that this interconnection among several qubits, termed 'quantum entanglement,' is essential to perform quantum computing.”

    From Quantum Computing in the Next-Generation Computational Biology Landscape: From Protein Folding to Molecular Dynamics (2024) · cites this paper as [17]

    Section 2.1.1 calls entanglement fundamental for *useful* quantum computation. It says algorithms without entanglement can be enacted on a classical computer with no significant speed difference; that supports its importance for quantum advantage, but not the unqualified claim that it is essential to perform quantum computing at all. (LLM, conf=0.85)

    Experimental — accuracy not yet measured

  • In recent years, the possibility of using quantum instead of or in combination with classical computing to solve biocomputing problems has emerged.

    “Interestingly, in recent years the possibility of using quantum instead of or in combination with classical computing to solve biocomputing problems has emerged [46, 47].”

    From Quantum computing algorithms: getting closer to critical problems in computational biology (2022) · cites this paper as [47]

The citation lists haven’t been fetched yet

References 163 · Cited by 232

Research works printed but not linked: 29

  1. Manin Y, 1980, Вычислимое и невычислимое (Computable and uncomputable)
  2. Szabo A, 2012, Modern quantum chemistry: Introduction to advanced electronic structure theory
  3. KitaevAY. Quantum measurements and the abelian stabilizer problem.arXiv preprint quant‐ph/9511026;1995. · arXiv
  4. Bi‐Xue W, 2018, Efficient quantum simulation of photosynthetic light harvesting, NPJ Quant Inform, 4, 1
  5. EmaniPS WarrellJ AnticevicA et al. Quantum computing at the frontiers of biological sciences.arXiv preprint arXiv:1911.07127;2019. · arXiv
  6. GottesmanD. The Heisenberg representation of quantum computers arXiv preprint quant‐ph/9807006;1998. · arXiv
  7. Li Y, 2017, Efficient variational quantum simulator incorporating active error minimization, Phys Rev X, 7, 021050
  8. DuttaS SuauA DuttaS RoyS BeheraBK PanigrahiBK. Demonstration of a quantum circuit design methodology for multiple regression.arXiv preprint arXiv:1811.01726;2018. · arXiv
  9. ChatterjeeR YuT. Generalized coherent states reproducing kernels and quantum support vector machines.arXiv preprint arXiv:1612.03713;2016. · arXiv
  10. WiebeB GranadeC. Can small quantum systems learn? arXiv preprint arXiv:1512.03145;2015. · arXiv
  11. Benedetti M, 2017, Quantum‐assisted learning of hardware‐embedded probabilistic graphical models, Phys Rev X, 7, 041052
  12. Kerenidis I, 2019, Advances in Neural Information Processing Systems, 4136
  13. KerenidisI LuongoA PrakashA. Quantum expectation‐maximization for Gaussian mixture models.arXiv preprint arXiv:1908.06657.2019.
  14. MiyaharaH AiharaK LechnerW. Quantum expectation–maximization algorithm.arXiv preprint arXiv:1908.06655;2019. · arXiv
  15. Kak SC, 1995, Advances in imaging and electron physics, 259
  16. Denil M., 2011, Toward the implementation of a quantum RBM
  17. Lippard SJ, 1994, Principles of bioinorganic chemistry
  18. ElfvingVE GámezJA GogolinC. Simulating quantum chemistry in the restricted Hartree–Fock space on a qubit‐based quantum computing device.arXiv preprint arXiv:2002.00035;2020 · arXiv
  19. GoogleAI Quantum and Collaborators. Hartree–Fock on a superconducting qubit quantum computer.arXiv preprint arXiv:2004.04174;2020. · arXiv
  20. Helgaker T, 2014, Molecular electronic‐structure theory
  21. Bravyi SB, 2002, Fermionic quantum computation, Ann Phys Rehabil Med, 298, 210
  22. BiamonteJ. Universal variational quantum computation.arXiv preprint arXiv:1903.04500;2019. · arXiv
  23. Messiah A, 2000, Quantum mechanics
  24. D‐Wave Systems Inc., 2019, D‐Wave problem‐solving handbook
  25. D‐Wave Systems Inc., 2019, Technical description of the D‐Wave quantum processing unit
  26. GoogleAI Quantum and Collaborators. Quantum approximate optimization of non‐planar graph problems on a planar superconducting processor.arXiv preprint arXiv:2004.04197;2020. · arXiv
  27. RobertS BarkoutsosPK WoernerS TavernelliI. Resource‐efficient quantum algorithm for protein folding.arXiv preprint arXiv:1908.02163;2019. · arXiv
  28. OuteiralC MorrisGM ShiJ StrahmM BenjaminSC DeaneCM. Investigating the potential for a limited quantum speedup on protein lattice problems.arXiv preprint arXiv:2004.01118;2020 · arXiv
  29. Rohl CA, 2004, Methods in enzymology, 66

Authors
Outeiral, C., Strahm, M., Shi, J., Morris, G. M., Benjamin, S. C., & Deane, C. M.
Affiliations
  • University of Oxford (GB)
  • F. Hoffmann-La Roche AG (CH)
  • UCB Pharma (United Kingdom) (GB)
Corresponding author
Deane, Charlotte M.
ORCID
Venue
WIREs Computational Molecular Science
Publisher
Wiley
Published
2020-05-22
Volume
11
Issue
1
ISSN
  • 1759-0876 (ISSN-L)
  • 1759-0884
DOI
10.1002/wcms.1481
OpenAlex
W3028293372
Semantic Scholar
3d82552eb483e5ea84b577a0e8d5f157a6085824
Cited by
232
References
163
Citations by year
2026: 31 · 2025: 52 · 2024: 45 · 2023: 50 · 2022: 27 · 2021: 25 · 2020: 2
FWCI
10.13
Type
journal article
Language
en
Open access
Licence
Locations
  • Wiley Interdisciplinary Reviews Computational Molecular Science (publishedVersion, cc-by)
  • arXiv (Cornell University) (submittedVersion)
  • arXiv (Cornell University) (cc-by)
Funders
Engineering and Physical Sciences Research Council (EP/L016044/1, EP/M013243/1, EP/S024093/1, EP/T001062/1)
Topics
  • Physical Sciences › Computer Science › Artificial Intelligence › Quantum Computing Algorithms and Architecture
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › DNA and Biological Computing
  • Physical Sciences › Computer Science › Computational Theory and Mathematics › Computational Drug Discovery Methods
Keywords
quantum computing, quantum algorithms, computational molecular biology, bioinformatics, quantum simulation, drug discovery, protein structure prediction, optimization algorithms, machine learning, network analysis, quantum advantage

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Claims read by deepseek-v4.1-flash under definition version 3 (extract-v3) on 2026-10-10.

Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-10-10.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review