Feedback-Based Quantum Control for Safe and Synergistic Drug Combination Design

Lan Nguyen Tran; Le Bin Ho; Mai Nguyen Phuong Nhi

REVIEW 4 major objections 5 minor 41 references

Feedback-Based Quantum Control for Safe and Synergistic Drug Combination Design

T0 review · 4 major / 5 minor · reviewed 2026-08-03 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read This paper shows that drug-combination safety and synergy can be encoded as an Ising Hamiltonian whose ground state is the optimal regimen, and that a feedback-based quantum algorithm called ITE-FALQON can find that ground state without cla

desk verdict Algebra is fine, but the main numerical result contradicts the paper's own Hamiltonian; the clinical findings are restatements of hand-assigned weights. read the letter →

arxiv 2601.18082 v1 pith:TDNGV7PF submitted 2026-01-26 quant-ph physics.chem-phphysics.comp-phphysics.med-ph

Mai Nguyen Phuong Nhi , Lan Nguyen Tran , Le Bin Ho This is my paper

classification quant-phphysics.chem-phphysics.comp-phphysics.med-ph

keywords drug-druginteractionsIsingHamiltonianFALQONimaginary-timeevolutionQUBOmaximumsafesubsetdrugcombinationoptimizationCOVID-19

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper's central claim is that selecting a safe and synergistic drug combination is a combinatorial optimization problem that can be written exactly as finding the ground state of an Ising Hamiltonian. Harmful drug pairs contribute penalties, synergistic pairs contribute rewards, and a cardinality term enforces regimen size. The authors then run a feedback-based quantum control algorithm, FALQON, and its imaginary-time extension ITE-FALQON, and show in simulation that these reach the optimal bitstrings for a six-drug network and a nine-drug COVID-19 panel. A sympathetic reader would care because the method removes the classical parameter-optimization loop that limits variational quantum algorithms, and because sparse, modular interaction graphs suggest the approach may scale to larger drug libraries. The clinical value of the results, however, depends entirely on the reliability of the hand-assigned interaction weights.

What carries the argument

The central object is the Ising Hamiltonian derived from the drug interaction graph, with decision variables z_i taking values in {+1, -1}. Harmful edges produce positive ZZ couplings that penalize co-selection, synergistic edges produce negative ZZ couplings that reward co-selection, and the cardinality constraint generates all-to-all couplings. The workhorse algorithm is FALQON, whose control field β(t) = -i⟨ψ(t)|[Hd, Hp]|ψ(t)⟩ is updated from the instantaneous commutator, implementing feedback-based gradient descent without any classical optimization loop. ITE-FALQON adds imaginary-time steps e^{-Δτ Hp} that exponentially suppress excited states. Together they convert clinical interaction

What would settle it

Run an exhaustive enumeration of all 64 subsets for the six-drug network with the Table I weights and verify that the two reported bitstrings are exactly the maximum safe subsets; likewise enumerate all 512 subsets for the COVID-19 panel and check that the reported triple is the unique SCO ground state for K=3. If exhaustive search returns a different optimum, the algorithm's output is not the true ground state. Independently, re-derive the interaction weights from the cited clinical databases; if the published optimal subsets change, the clinical conclusions are weight artifacts.

Watch

Extended reading notes

Core claim

The authors establish that the Maximum Safe Subset problem maps to an Ising Hamiltonian with purely penalizing couplings, so its ground state is exactly the largest harm-free drug set, and that the Synergy-Constrained Optimization problem, which adds synergy rewards and a quadratic cardinality term, has as its ground state the optimal clinically constrained regimen. They further demonstrate numerically that ITE-FALQON, which alternates feedback-controlled unitary evolution with imaginary-time filtering, converges to these ground states for all tested penalty parameters, while standard FALQON approaches but does not fully reach them. In the COVID-19 case study, the K=3 solution returns the tr

Load-bearing premise

The clinical validity of the interaction weights in Tables I and III: the paper describes them as inferred from clinical severity or from literature, but gives no reproducible mapping from source data to the normalized numbers, and every reported optimal regimen is fully determined by these weights.

Editorial extensions

If this is right

Any DDI database expressible as pairwise interaction weights can be plugged into the same Hamiltonian construction, making the framework a general recipe for turning interaction tables into optimization circuits.
ITE-FALQON's exact convergence to the MSS ground states in simulation means the remaining bottleneck is not the solver but the quality of the encoded weights.
Because the SCO Hamiltonian encodes regimen size as a soft constraint, clinicians can explore any target size K by changing one parameter, and the solver returns a single dominant solution.
The COVID-19 example indicates the method can recover experimentally supported combinations from pairwise data alone, without higher-order interaction terms.
Sparsity and modularity of real DDI graphs imply the approach may extend to roughly 50-100 qubits on near-term hardware, given appropriate preprocessing.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

The encoding is exact, so the same Hamiltonian could be handed to any ground-state solver — classical simulated annealing, QAOA, or quantum annealing — and the paper's specific contribution is showing that a parameter-free feedback controller is one viable solver; the framework does not require quantum hardware to be useful as a formulation.
A natural stress test is to replace the hand-normalized weights with independently extracted values from structured clinical databases and check whether the optimal subsets change; if they do, the method's output is weight-sensitive and future work should treat weights as uncertain inputs.
Because the MSS problem with penalties is the weighted maximum independent set on the harmful graph, connecting to known MIS results would allow a direct comparison of FALQON against established classical heuristics on much larger graphs.
The soft quadratic cardinality term may admit solutions slightly off-target size for finite µ; a testable extension would harden the constraint by increasing µ adaptively or adding a constraint-preserving mixer.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Hi — quick read of 2601.18082. The algebraic core is fine: the MSS and SCO loss functions correctly map to Ising Hamiltonians, and FALQON/ITE-FALQON are real, sensible optimizers. But the central numerical claim doesn't hold up as written. Under the natural ordering of the six drugs in Sec. IV.A, the reported MSS bitstrings 111000 and 110100 both contain harmful pairs — Ritonavir-Everolimus and Ritonavir-Cabazitaxel for the first; Ritonavir-Everolimus, Ritonavir-Metformin, and Everolimus-Metformin for the second. Both incur positive penalties in Eq. (3), so neither can be a ground state for any α>0. The paper never defines a bit-to-drug mapping, so the reader can't verify the result; the stress-test note is right on the mark.

What the paper does well: the derivation is clean, the reduction to QUBO is standard but correct, and using FALQON/ITE-FALQON as off-the-shelf optimizers is a reasonable proof-of-principle. The COVID case study is a legitimate illustration of the workflow, even if small.

The soft spots are real. The clinical weights in Tables I and III are hand-inferred from sources like Drugs.com without a reproducible mapping, and the COVID K=3 triple {RDV, RBV, MOV} is exactly the highest-synergy triple in Table III, so calling it a 'recovered clinically meaningful combination' is circular. There are no baselines, no error bars, no code verification, and n=6/n=9 are trivially solvable classically. The QUBO formulations themselves are covered by the paper's own ref [31], so novelty is thin.

In sum: this is a proof-of-principle with a correct framework but a contradictory demonstration. The bitstring issue may be a simple typo or an unstated ordering choice, but the paper doesn't give the reader enough to check. A serious referee could help pin that down and demand reproducible data and baselines. I'd send it out only if the editor expects a major revision; in current form it should not pass peer review.

Recommendation: don't desk-reject outright if the authors can clarify the mapping and rerun with transparent provenance — but as it stands, the paper is not publishable.

Referee Report

4 major / 5 minor

Summary. The paper proposes a feedback-based quantum control framework (FALQON and its imaginary-time extension ITE-FALQON) for optimizing drug combinations subject to drug-drug interaction constraints. It defines two QUBO/Ising formulations: the Maximum Safe Subset (MSS) problem, which rewards drug inclusion and penalizes harmful pairs, and the Synergy-Constrained Optimization (SCO) problem, which adds synergy rewards and a cardinality penalty. The authors derive the corresponding Ising Hamiltonians, simulate the algorithms on a six-drug network using weights stated in Table I, and apply the method to a nine-drug COVID-19 case study. They report that ITE-FALQON converges to ground states and that the recovered combinations are clinically meaningful. The algebraic mappings in Eqs. (5)-(7) and (10)-(12) are correct. However, the central numerical results suffer from an internal inconsistency in the reported bitstring-to-drug mapping, and the clinical conclusions are largely predetermined by the hand-assigned interaction weights.

Significance. If the claims were fully supported, the paper would offer a useful proof-of-concept that feedback-based quantum optimization can be applied to DDI-encoded QUBOs without classical parameter optimization. The algebra of the Ising encodings is correct, and the authors' choice to make the algorithm feedback-driven rather than variational is well motivated. The manuscript also provides a data-availability link, which is a positive step for reproducibility. That said, the current demonstration is limited to tiny instances (n=6 and n=9) that are trivially solvable by exhaustive search, and the main quantitative results appear to be inconsistent with the stated variable mapping. The clinical 'recovery' results are also direct consequences of the authors' own weight assignments, not independent validation. The paper's value therefore hinges on correcting the numerical inconsistencies and clarifying the role of the hand-assigned weights.

major comments (4)

[§IV.B and Fig. 3] Under the drug ordering introduced in Eq. (1) and used in Sec. IV.A ('Ritonavir, Everolimus, Cabazitaxel, Metformin, Erlotinib, Topotecan'), the reported dominant MSS bitstrings cannot be ground states of Eq. (3). The string 111000 corresponds to {Ritonavir, Everolimus, Cabazitaxel}, which contains the harmful pairs (Ritonavir, Everolimus) and (Ritonavir, Cabazitaxel) with weights 0.95 and 0.90 in Table I; 110100 corresponds to {Ritonavir, Everolimus, Metformin}, which contains the harmful pairs (Ritonavir, Everolimus), (Ritonavir, Metformin), and (Everolimus, Metformin). For any alpha>0, both have L_MSS = -3 + c*alpha with c>0, while the harm-free subset {Metformin, Erlotinib, Topotecan} gives L_MSS = -3, so these strings are strictly worse than valid configurations of the same cardinality. The statement that 'both configurations exclude all major harmful interactions' is therefore dire
[§II.A and §V.A] The claim that 'the ground state of this Hamiltonian corresponds exactly to the largest safe drug subset' is only true for sufficiently large alpha. Eq. (3) is a weighted trade-off: a harmful pair will be accepted if the cardinality reward outweighs the penalty. This is not merely a theoretical caveat; the paper's own COVID-19 MSS result at alpha=2.5 in Sec. V.A explicitly retains the harmful PAX-DEX pair. The exactness claim should be restricted to alpha above a proved threshold, or the MSS solution should be described as a risk-tolerance-dependent optimum rather than the unique maximum safe subset.
[§II, Tables I and III] The clinical input data are not reproducible. The text states that weights are 'extracted from established pharmacological resources' and normalized, but Table I is captioned 'weights inferred from clinical severity' and Table III says 'weights from literature' without per-edge citations. No mapping from any source to any numerical weight is given. Since every reported optimum is fully determined by these numbers, the clinical conclusions cannot be independently checked. A reproducible protocol—with source, extraction rule, and normalization—is required before the results can be assessed as clinically meaningful.
[§V.B and Table III] The COVID-19 SCO result {RDV, RBV, MOV} is largely a restatement of the input weights. In Table III, this triple contains the three largest synergy entries (0.95, 0.80, 0.60) and no harmful edges, so it is the expected optimum of Eq. (8) for essentially any reasonable choice of parameters. The paper should explicitly frame this as a consistency check of the encoding, not as 'recovering clinically meaningful combinations' or as independent evidence that the quantum algorithm discovers new clinical knowledge.

minor comments (5)

[§III.A, Eq. (13)] The feedback law is written as β(t) = -i⟨[H_d, H_p]⟩, which differs in sign convention from Ref. [35] where the commutator is usually written as -i⟨[H_p, H_d]⟩ or with an explicit imaginary prefactor. Please reconcile the sign convention to avoid implying energy ascent.
[§IV.C] The sentence 'why we fix γ=2.5 and μ=5' should read 'where we fix γ=2.5 and μ=5'.
[§V.A] Several grammatical errors appear: 'which no harmful interaction at all' should be 'which has no harmful interactions at all'; 'the optimal solutions is' should be 'the optimal solution is'.
[§II and Table II] There are typos and inconsistencies in names: 'SYNERYxDB.ca' should be 'SYNERGxDB'; 'Mpr o/3CLpro' should be 'Mpro/3CLpro'; reference [36] is truncated ('for univ').
[§IV, Figs. 2 and 4] The exact ground-state energies are not shown in the energy evolution plots; adding horizontal dashed lines or numeric energy values would make the claimed convergence to the ground state verifiable. Also, please provide a table or explicit mapping from bitstrings to drug names for every reported dominant configuration.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the Ising encodings are explicit and the reported optima are direct optimizers of the stated objectives.

full rationale

The paper's derivation chain is an explicit QUBO-to-Ising encoding: Eq. (3) defines the MSS loss with selection rewards and harmful-pair penalties, and Eqs. (5)-(7) give the Ising coefficients by the standard xi=(1-zi)/2 substitution. The ground-state correspondence is thus a theorem of the stated objective for sufficiently large α, not a prediction fitted from data. Similarly, the SCO loss (Eq. 8) defines its own objective, and the reported K=3 COVID triple {RDV, RBV, MOV} is the output of that objective for the assigned Table III weights; being determined by the input weights is the intended behavior of an optimizer, not a circular reduction. The only self-citation is [36] for the ITE-FALQON method; the method is described in situ and its convergence is demonstrated by the paper's own statevector simulations (Figs. 2, 4, 6), so the citation is attribution rather than a load-bearing circular justification. The manuscript does raise separate evidence concerns: the interaction-weight provenance is not fully reproducible (Tables I and III are said to be 'inferred' or 'from literature' with no mapping), and the reported MSS bitstrings 111000 and 110100 appear inconsistent with Table I under the paper's natural bit ordering. These are correctness/reproducibility problems, not equivalences of outputs to inputs by construction, so they do not raise the circularity score.

Assumptions & free parameters 7 free parameters · 5 assumptions · 0 invented entities

The central claim rests on: (i) the standard Ising/QUBO substitution (textbook), (ii) domain assumptions that the cited DDI databases are accurate and that pairwise interactions suffice, (iii) the assumption that hand-normalized weights reflect clinical severity/synergy, and (iv) the convergence behavior of FALQON (external) and ITE-FALQON (the authors' prior work, not independently verified here). The key free parameters are the interaction weights themselves plus alpha, gamma, mu, K, and Delta-tau.

free parameters (7)

interaction weights w_ij (harm/synergy) = Tabulated values, e.g. Ritonavir-Everolimus harm 0.95, RBV-RDV synergy 0.95
Hand-inferred from clinical severity (Table I caption) or normalized from literature (Table III); every reported optimal subset is a function of these values.
MSS penalty coefficient alpha = 2.5 and 5
Chosen to encode clinical risk tolerance; the paper concludes 'larger alpha is better,' i.e. the qualitative result depends on the chosen value.
SCO harm penalty gamma = 2.5
Fixed ('why we fix gamma = 2.5 and mu = 5', Sec. IV.C) without sensitivity analysis.
SCO cardinality penalty mu = 5
Hand-chosen; controls whether the K constraint is actually enforced in the ground state.
SCO target cardinality K = 3 and 4
User-set design parameter that shifts the optimal solution and energy.
imaginary-time step Delta-tau = 0.1 (MSS), 0.01 (SCO)
Chosen 'to ensure numerical stability'; affects convergence behavior of ITE-FALQON.
evolution depth T = unspecified
Number of feedback steps is never stated, so the reported convergence curves cannot be reproduced exactly.

assumptions (5)

standard math The substitution xi=(1-zi)/2 and the coefficient expressions in Eqs. (5)-(7) and (10)-(12) faithfully encode the loss functions (3) and (8).
Verified by hand; standard QUBO-to-Ising mapping (cf. ref [31]).
domain assumption Drug-drug interaction data from Drugs.com, SYNERGxDB, and the Liverpool COVID-19 database are accurate and sufficient; higher-order (non-pairwise) interactions are negligible.
Invoked in Sec. II and stressed in Sec. V ('relies only on pairwise interactions'). Real polypharmacy risks include higher-order effects, acknowledged only qualitatively in Sec. V.
domain assumption Normalized weights in [0,1] reflect relative clinical severity and synergy strength.
The normalization scheme is asserted (Sec. II, Table I caption), never derived or validated against an external scale.
standard math The discretized FALQON feedback rule beta(t) = -i<[Hd,Hp]> drives the objective energy monotonically downward toward the ground state.
Continuous-time property from ref [35]; the paper relies on it for discretized evolution (Sec. III.A).
ad hoc to paper ITE-FALQON as described in ref [36] reaches the exact ground state with the stated Delta-tau choices.
Self-cited prior work (co-authors Tran, Ho); convergence is demonstrated only by the paper's own figures, not independently verified or proven here.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Feedback-Based Quantum Control for Safe and Synergistic Drug Combination Design." pith.science (2026). https://pith.science/paper/TDNGV7PF

@misc{pith2026260118082,
  author       = {Pith},
  title        = {Pith review of: Feedback-Based Quantum Control for Safe and Synergistic Drug Combination Design},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/TDNGV7PF}},
  note         = {Machine review of arXiv:2601.18082}
}

read the original abstract

Drug-drug interactions (DDIs) strongly affect the safety and efficacy of combination therapies. Despite the availability of large DDI databases, selecting optimal multi-drug combinations that balance safety, therapeutic benefit, and regimen size remains a challenging combinatorial optimization problem. Here, we present a quantum-control-based framework for DDI-aware drug combination optimization, in which known harmful and synergistic interactions are encoded into Ising Hamiltonians as penalties and rewards, respectively. The optimization is performed using the feedback-based quantum algorithm FALQON, a gradient-free variational approach. We study two clinically motivated tasks: the Maximum Safe Subset problem and the Synergy-Constrained Optimization problem. Numerical simulations using interaction data from Drugs.com and SYNERGxDB demonstrate efficient convergence and high-quality solutions for clinically relevant drug sets, including COVID-19 case studies.

Figures

Figures reproduced from arXiv: 2601.18082 by the authors.

**Figure 2.** FIG. 2. Energy evolution obtained using FALQON and ITE [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗

**Figure 3.** FIG. 3. Final probability distributions and dominant MSS so [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗

Figures from the paper (3 more)

**Figure 4.** Figure 4: FIG. 4. SCO results obtained using ITE-FALQON. (a) Energy co [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]

**Figure 5.** Figure 5: FIG. 5. Interaction graph of the 9-drug network. [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]

**Figure 7.** Figure 7: FIG. 7. (a) Energy evolution obtained from ITE-FALQON [PITH_FULL_IMAGE:figures/full_fig_p008_7.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

41 extracted references · 6 canonical work pages

[31]

J. Chen, A. Lin, A. Jiang, C. Qi, Z. Liu, Q. Cheng, S. Yuan, and P. Luo, Computational frameworks trans- form antagonism to synergy in optimizing combination therapies, npj Digital Medicine 8, 44 (2025)

2025
[1]

N. Abdu, T. Tewelde, N. Jabir, S. Idrisnur, and E. H. Tesfamariam, Evaluation of potential drug– drug interactions and its determinants among out- patient prescriptions in six community chain phar- macies in asmara, eritrea: a cross-sectional study, Scientiﬁc Reports 15, 33403 (2025)

2025
[2]

Con- sequently, the ground state of HSCO encodes the optimal clinically constrained drug set, simultaneously balancing eﬃcacy, safety, and size

(12) The sign and magnitude of the couplings have clear pharmacological meaning: synergistic pairs generate negative couplings, which energetically favor their co- selection, harmful pairs generate positive couplings, which penalize conﬁgurations containing both drugs, and the cardinality term induces global correlations that steer the system toward the p...

2024
[3]

L. Li, J. Baker, R. Quirk, D. Deidun, M. Moran, A. A. Salem, N. Aryal, B. A. Van Dort, W. Y. Zheng, A. Hargreaves, P. Doherty, S. N. Hilmer, R. O. Day, J. I. Westbrook, and M. T. Baysari, Drug–drug in- teractions and actual harm to hospitalized patients: A multicentre study examining the prevalence pre- and post-electronic medication system implementation...

2024
[4]

W. Y. Zheng, L. C. Richardson, L. Li, R. O. Day, J. I. Westbrook, and M. T. Baysari, Drug-drug in- teractions and their harmful eﬀects in hospitalised patients: a systematic review and meta-analysis, European Journal of Clinical Pharmacology 74, 15 (2018)

2018
[5]

Buc¸ sa, A

C. Buc¸ sa, A. Farca¸ s, I. Cazacu, D. Leucuta, A. Achimas-Cadariu, C. Mogosan, and M. Bojita, How many potential drug–drug interactions cause adverse drug reactions in hospitalized patients?, European Journal of Internal Medicine 24, 27 (2013)

2013
[6]

Novak, A

J. Novak, A. Goldberg, K. Dharmarajan, A. Amini, R. J. Maggiore, C. J. Presley, and G. Nightingale, Polypharmacy in older adults with cancer undergoing radiotherapy: A review, Journal of Geriatric Oncology 13, 778 (2022)

2022
[7]

Schmitz, C

A. Schmitz, C. Reichardt, J. Gierth, M. Richter, F. Mrosk, M. Alfertshofer, L. Knoedler, C. Doll, F. Haus- mann, M. Lee, B. Kleikamp, C. Rendenbach, M. Heiland, and S. Koerdt, Retrospective analysis of multimorbid- ity, polypharmacy, and drug interactions on postopera- tive outcomes in oral squamous cell carcinoma patients, Clinical Oral Investigations 29...

2025

Show all 41 references

[8]

R. M. Alhumaidi, G. A. Bamagous, S. M. Alsanosi, H. S. Alqashqari, R. S. Qadhi, Y. Z. Alhindi, N. Ayoub, and A. H. Falemban, Risk of polypharmacy and its outcome in terms of drug interaction in an elderly population: A retrospective cross-sectional study, Journal of Clinica l ...

2023 doi
[9]

M. R. Mohamed, S. G. Mohile, K. M. Juba, H. Awad, M. Wells, K. P. Loh, M. Flannery, E. Culakova, R. G. Tylock, and E. E. Ramsdale, Association of polypharmacy and potential drug-drug inter- actions with adverse treatment outcomes in older adults with advanced cancer, Cancer 12...

2023 doi
[10]

Ballestri, E

S. Ballestri, E. Romagnoli, D. Arioli, V. Coluc- cio, A. Marrazzo, A. Athanasiou, M. Di Girolamo, C. Cappi, M. Marietta, and M. Capitelli, Risk and management of bleeding complications with direct oral anticoagulants in patients with atrial ﬁbrillation and venous thromboemboli...

2023
[11]

Donz´ e, N

J. Donz´ e, N. Rodondi, G. Waeber, J. Cornuz, and D. Au- jesky, Major bleeding risk in anticoagulated patients re- ceiving concomitant antiplatelet therapy: A prospective study, Thrombosis Research 131, 502 (2013)

2013
[12]

Miernik, A

S. Miernik, A. Matusiewicz, and M. Olesi´ nska, Drug- induced myopathies: A comprehensive review and up- date, Biomedicines 12, 10.3390/biomedicines12050987 (2024)

2024 doi
[13]

Janssen, N

L. Janssen, N. A. E. Allard, C. G. J. Saris, J. Kei- jer, M. T. E. Hopman, and S. Timmers, Muscle toxicity of drugs: When drugs turn physiology into pathophys- iology, Physiological Reviews 100, 633 (2020), pMID: 31751166, https://doi.org/10.1152/physrev.00002.2019

2020
[14]

C. R. Dormuth, B. R. Hemmelgarn, J. M. Pa- terson, M. T. James, G. F. Teare, C. B. Ray- mond, J.-P. Lafrance, A. Levy, A. X. Garg, and P. Ernst, Use of high potency statins and rates of admission for acute kidney injury: multicen- ter, retrospective observational analysis of a...

2013 doi
[15]

Duarte and N

D. Duarte and N. Vale, Evaluation of syn- ergism in drug combinations and reference models for future orientations in oncology, Current Research in Pharmacology and Drug Discovery 3, 100110 (2022)

2022
[16]

S. K. Singh, A. Mohammed, O. A. Alghamdi, and S. M. Husain, Chapter 12 - new approaches for targeting drug resistance through drug combination, in Combination Therapy Against Multidrug Resistance , edited by M. Y. Wani and A. Ahmad (Academic Press,
[17]

Z. A. Shyr, Y.-S. Cheng, D. C. Lo, and W. Zheng, Drug combination therapy for emerging viral diseases, Drug Discovery Today 26, 2367 (2021)

2021
[18]

Calzetta, C

L. Calzetta, C. Page, M. G. Matera, M. Cazzola, and P. Rogliani, Drug-drug interactions and synergy: From pharmacological models to clinical application, Pharmacological Reviews 76, 1159 (2024)

2024
[19]

Vanessa, T

P. Vanessa, T. Nina, J. J. M., W. Brian, and K. F. C., Combinatorial approaches to the prevention and treatment of hiv-1 infection, Antimicrobial Agents and Chemotherapy 55, 1831 (2011)

2011
[20]

R. B. Mokhtari, T. S. Homayouni, N. Baluch, E. Morgatskaya, S. Kumar, B. Das, and H. Yeger, Combination therapy in combating cancer, Oncotarget 8, 38022 (2017)

2017
[21]

R. J. Worthington and C. Melander, Combination approaches to combat multidrug-resistant bacteria, Trends in Biotechnology 31, 177 (2013), special Issue: Celebrating 30 years of biotechnology

2013
[22]

Akinbolade, D

S. Akinbolade, D. Coughlan, R. Fairbairn, G. McConkey, H. Powell, D. Ogunbayo, and D. Craig, Combination therapies for covid-19: An overview of the clinical trials landscape, British Journal of Clinical Pharmacology 88, 1590 (2022), https://bpspubs.onlinelibrary.wiley.com/doi/...

2022 doi
[23]

Bobrowski, L

T. Bobrowski, L. Chen, R. T. Eastman, Z. Itkin, P. Shinn, C. Z. Chen, H. Guo, W. Zheng, S. Michael, A. Simeonov, M. D. Hall, A. V. Zakharov, and E. N. Muratov, Synergistic and antagonistic drug combinations against sars-cov-2, Molecular Therapy 29, 873 (2021)

2021
[24]

G. Li, R. Hilgenfeld, R. Whitley, and E. De Clercq, Therapeutic strategies for covid-19: progress and lessons learned, Nature Reviews Drug Discovery 22, 449 (2023)

2023
[25]

Y. Wang, Z. Yang, and Q. Yao, Accurate and interpretable drug-drug interaction predic- tion enabled by knowledge subgraph learning, Communications Medicine 4, 59 (2024)

2024
[26]

S. Wang, H. Yuan, Z. Hong, X.-g. Chen, and X. Yang, Path-based graph neural network for drug synergy pre- diction and interpretation, Journal of Chemical Informa- tion and Modeling 10.1021/acs.jcim.5c02569 (2025)

2025 doi
[27]

J. Y. Ryu, H. U. Kim, and S. Y. Lee, Deep learning improves prediction of drug–drug and drug–food interactions, Proceedings of the National Academy of Sciences 115, E4304 (2018), https://www.pnas.org/doi/pdf/10.1073/pnas.1803294115

2018 doi
[28]

Fillinger, S

L. Fillinger, S. Walter, M. Ley, K. Keska-Izworska, L. G . Dehkordi, K. Kratochwill, and P. Perco, Computational modeling approaches and regulatory pathways for drug combinations, Drug Discovery Today 30, 104345 (2025)

2025
[29]

H. Luo, W. Yin, J. Wang, G. Zhang, W. Liang, J. Luo, and C. Yan, Drug-drug interactions prediction based on deep learning and knowledge graph: A review, iScience 27, 109148 (2024)

2024
[30]

Zitnik, M

M. Zitnik, M. Agrawal, and J. Leskovec, Modeling polypharmacy side eﬀects with graph convolutional net- works, Bioinformatics 34, i457 (2018)

2018
[32]

Glover, G

F. Glover, G. Kochenberger, and Y. Du, A tutorial on formulating and using qubo models (2019), arXiv:1811.11538 [cs.DS]

2019 arXiv
[33]

Bittel and M

L. Bittel and M. Kliesch, Training vari- ational quantum algorithms is np-hard, Phys. Rev. Lett. 127, 120502 (2021)

2021
[34]

Farhi, J

E. Farhi, J. Goldstone, and S. Gutmann, A quantum approximate optimization algorithm (2014), arXiv:1411.4028 [quant-ph]

2014 arXiv
[35]

J. R. McClean, S. Boixo, V. N. Smelyanskiy, R. Babbush, and H. Neven, Barren plateaus in quantum neural network training landscapes, Nature Communications 9, 4812 (2018)

2018
[36]

A. B. Magann, K. M. Rudinger, M. D. Grace, and M. Sarovar, Feedback-based quantum optimization, Phys. Rev. Lett. 129, 250502 (2022)

2022
[37]

T. N. V. Long, L. N. Tran, and L. B. Ho, Imaginary-time-enhanced feedback-based quantum algorit hms for univ (2025), arXiv:2512.13044 [quant-ph]

2025
[38]

Chou, Theoretical basis, experimental de- sign, and computerized simulation of synergism and antagonism in drug combination studies, Pharmacological Reviews 58, 621 (2006)

T.-C. Chou, Theoretical basis, experimental de- sign, and computerized simulation of synergism and antagonism in drug combination studies, Pharmacological Reviews 58, 621 (2006)

2006
[39]

Rahmah, S

L. Rahmah, S. O. Abarikwu, A. G. Arero, M. Essouma, A. T. Jibril, A. Fal, R. Flisiak, R. Makuku, L. Mar- quez, K. Mohamed, L. Ndow, D. Zarkebska-Michaluk, N. Rezaei, and P. Rzymski, Oral antiviral treatments for 11 covid-19: opportunities and challenges, Pharmacol. Rep. 74, 12...

2022
[40]

D. K. Brady, A. R. Gurijala, L. Huang, A. A. Hussain, A. L. Lingan, O. G. Pembridge, B. A. Ratangee, T. T. Sealy, K. T. Vallone, and T. P. Clements, A guide to COVID-19 antiviral therapeutics: a summary and per- spective of the antiviral weapons against SARS-CoV-2 infection, F...

2024
[41]

M. N. P. Nhi, L. N. Tran, and L. B. Ho, Sup- porting data for drug-drug interaction optimiza- tion using the feedback-based quantum control al- gorithms (2026), https://github.com/echkon/ddi., https://github.com/echkon/ddi (2026)

2026

Pith tools

Reviewed August 3, 2026 · model on record in the stance chip above.

[1] [31]

J. Chen, A. Lin, A. Jiang, C. Qi, Z. Liu, Q. Cheng, S. Yuan, and P. Luo, Computational frameworks trans- form antagonism to synergy in optimizing combination therapies, npj Digital Medicine 8, 44 (2025)

2025

[2] [1]

N. Abdu, T. Tewelde, N. Jabir, S. Idrisnur, and E. H. Tesfamariam, Evaluation of potential drug– drug interactions and its determinants among out- patient prescriptions in six community chain phar- macies in asmara, eritrea: a cross-sectional study, Scientiﬁc Reports 15, 33403 (2025)

2025

[3] [2]

Con- sequently, the ground state of HSCO encodes the optimal clinically constrained drug set, simultaneously balancing eﬃcacy, safety, and size

(12) The sign and magnitude of the couplings have clear pharmacological meaning: synergistic pairs generate negative couplings, which energetically favor their co- selection, harmful pairs generate positive couplings, which penalize conﬁgurations containing both drugs, and the cardinality term induces global correlations that steer the system toward the p...

2024

[4] [3]

L. Li, J. Baker, R. Quirk, D. Deidun, M. Moran, A. A. Salem, N. Aryal, B. A. Van Dort, W. Y. Zheng, A. Hargreaves, P. Doherty, S. N. Hilmer, R. O. Day, J. I. Westbrook, and M. T. Baysari, Drug–drug in- teractions and actual harm to hospitalized patients: A multicentre study examining the prevalence pre- and post-electronic medication system implementation...

2024

[5] [4]

W. Y. Zheng, L. C. Richardson, L. Li, R. O. Day, J. I. Westbrook, and M. T. Baysari, Drug-drug in- teractions and their harmful eﬀects in hospitalised patients: a systematic review and meta-analysis, European Journal of Clinical Pharmacology 74, 15 (2018)

2018

[6] [5]

Buc¸ sa, A

C. Buc¸ sa, A. Farca¸ s, I. Cazacu, D. Leucuta, A. Achimas-Cadariu, C. Mogosan, and M. Bojita, How many potential drug–drug interactions cause adverse drug reactions in hospitalized patients?, European Journal of Internal Medicine 24, 27 (2013)

2013

[7] [6]

Novak, A

J. Novak, A. Goldberg, K. Dharmarajan, A. Amini, R. J. Maggiore, C. J. Presley, and G. Nightingale, Polypharmacy in older adults with cancer undergoing radiotherapy: A review, Journal of Geriatric Oncology 13, 778 (2022)

2022

[8] [7]

Schmitz, C

A. Schmitz, C. Reichardt, J. Gierth, M. Richter, F. Mrosk, M. Alfertshofer, L. Knoedler, C. Doll, F. Haus- mann, M. Lee, B. Kleikamp, C. Rendenbach, M. Heiland, and S. Koerdt, Retrospective analysis of multimorbid- ity, polypharmacy, and drug interactions on postopera- tive outcomes in oral squamous cell carcinoma patients, Clinical Oral Investigations 29...

2025