Pith. sign in

REVIEW 3 major objections 5 minor 53 references

Multi-Objective-Guided Discrete Flow Matching for Controllable Biological Sequence Design

T0 review · 3 major / 5 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read Sampling-time guidance turns any pretrained discrete flow matching model into a multi-objective sequence designer.

desk verdict MOG-DFM is a real algorithmic contribution for steering discrete flow samplers, but the formal guarantee is essentially definitional and the evaluation is mostly circular; send it to review, but expect major revisions. read the letter →

arxiv 2505.07086 v2 pith:VVERKBFB submitted 2025-05-11 cs.LG q-bio.BM

Tong Chen , Yinuo Zhang , Sophia Tang , Pranam Chatterjee This is my paper · ORCID
classification cs.LGq-bio.BM
keywords discreteflowmatchingmulti-objectiveoptimizationParetofrontpeptidebinderdesignenhancerDNAguidedgenerationcontrollablesequenceCTMCsampling
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

At its core, MOG-DFM is a sampling-time steering layer for discrete flow matching. Instead of retraining a generator, it reweights the model's token-level transition velocities so that, step by step, the partially generated sequence moves in a direction that improves several competing properties at once. The paper shows that, under its scoring and filtering scheme, each accepted token replacement has positive expected improvement along a chosen trade-off weight vector, which in expectation directs generation toward the Pareto front. This is demonstrated on therapeutic peptide design with five simultaneous objectives, and on enhancer DNA design with class and shape objectives, where the method outperforms classical multi-objective optimizers and flow baselines in producing balanced profiles.

What carries the argument

The load-bearing objects are the guided transition score and the adaptive hypercone filter. For each candidate token replacement $y_i$ at position $i$, the score is $\Delta S(y_i,x,\omega) = \mathrm{Norm}\left(\frac{1}{N}\sum_{n=1}^N i_n I_n(y_i,x)\right) + \lambda\, \mathrm{Norm}\left(\Delta s(y_i,x)\cdot\omega\right)$, where $I_n$ is the rank-normalized improvement of objective $n$ among all tokens at that position, $\Delta s$ is the vector of score changes, and $\omega$ is the sampled trade-off weight. The guided velocity is $\beta u_t^i(y_i,x)\exp(\Delta S)$, which preserves the non-negativity and zero-sum rate conditions, so the process remains a valid continuous-time Markov chain. The hypercone accepts only candidates with angle $\alpha_i = \arccos\left(\frac{\Delta s\cdot\omega}{\|\Delta s\|\|\omega\|}\right) \le \Phi_t$, and $\Phi_t$ adapts via an exponential moving average of the rejection rate, widening when too many candidates are rejected and narrowing when too few are. Together, these mechanisms bias the dynamics toward locally improving, directionally aligned moves and prevent the trajectory from drifting off the chosen trade-off direction.

What would settle it

Run the full pipeline with the property scorers replaced by random noise; if the guided sequences still show large balanced gains on the original predictors, the apparent Pareto improvement is an artifact of the guidance loop rather than real optimization. A direct lab assay of designed binders, measuring hemolysis, solubility, half-life, and affinity, would also settle the question, because predictor-only gains would fail to reproduce in measurement.

Watch

Extended reading notes

Core claim

The paper claims that multi-objective control of discrete flow matching can be achieved purely at sampling time. The generator is treated as a black box; a precomputed set of weight vectors is sampled, and at each sampling step one position in the sequence is chosen and every alternative token at that position is scored. The score combines a rank-normalized per-objective improvement and a directional term that measures how well the improvement vector aligns with the sampled weight vector. The original transition velocity is multiplied by the exponent of this score, while keeping the non-negativity and zero-sum rate conditions intact. Then a hypercone filter retains only those token replacements whose improvement vector lies within an angular tolerance of the weight vector, and that tolerance adapts based on how many candidates are rejected. The paper proves that, if the candidate set is nonempty, the expected improvement along the weight vector is strictly positive. On peptide binders, the method yields simultaneous gains in hemolysis reduction, non-fouling, solubility, half-life, and affinity across multiple targets; on enhancer DNA, it achieves the requested class and shape values, and ablations show that removing any guidance criterion collapses that property.

Load-bearing premise

The quantitative claims assume that the machine-learning predictors used both to guide and to evaluate the sequences faithfully reflect real biological properties; if those predictors are biased, the reported Pareto improvements could be artifacts of the score functions rather than genuine gains in molecule quality.

Editorial extensions

If this is right

  • Any pretrained discrete flow matching model over a finite vocabulary can be steered toward multi-objective trade-offs without retraining, by reweighting its token-level velocities at sampling time.
  • User-supplied weight vectors from a Das-Dennis simplex lattice let separate runs explore different regions of the Pareto front, so a single base generator can cover diverse trade-off preferences.
  • In peptide binder design, simultaneous guidance across hemolysis, non-fouling, solubility, half-life, and binding affinity moves all five scores in the improving direction over 100 iterations, with gains corroborated by an independent ADMET predictor and by docking and structure-prediction tools.
  • In enhancer DNA design, the same framework steers sequences toward a specified enhancer class and DNA shape value (HelT or Rise), and ablations show that removing a guidance criterion collapses that property.
  • The hypercone ablation shows that removing the filter collapses half-life gains from roughly 30-35 hours down to 4-13 hours, indicating that angular filtering is what allows gradual, coordinated objectives to progress.

Reading between the lines

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

  • Editorial inference: because the guidance only touches the velocity field and requires only scalar scorers, the same recipe should transfer to other discrete sequence domains such as RNA design, antibody CDR optimization, or promoter design without modifying the base model, though the paper does not demonstrate this.
  • Editorial inference: the proof establishes positive expected improvement along a single weight direction, but it does not establish a convergence rate to the full Pareto set; whether many runs with different weight vectors cover the front depends on the predictor landscape, so a formal coverage guarantee would be a natural next step.
  • Editorial inference: the paper cross-checks solubility and half-life with an independent predictor and binding with structure tools, but hemolysis and non-fouling rest on the original classifiers alone; a wet-lab assay or an additional orthogonal predictor for those two properties would strengthen the claim.
  • Editorial inference: because rank normalization amplifies small score differences, high-noise predictors such as the hemolysis classifier (F1 $=0.58$) could inject noise into the guidance signal; ensembling predictors or making the rank term uncertainty-aware is a plausible robustness extension the paper does not test.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 5 minor

Summary. The paper introduces MOG-DFM, a guidance framework for pretrained discrete flow matching models that aims to generate biological sequences satisfying multiple, possibly conflicting objectives. At each sampling step, MOG-DFM scores candidate token replacements with a hybrid rank-directional score, reweights the base model's transition rates, applies an adaptive hypercone filter to restrict moves to those roughly aligned with a user-selected trade-off vector, and then evolves the CTMC by Euler sampling. The authors train two unconditional base models, PepDFM for peptides and EnhancerDFM for enhancer DNA, and evaluate MOG-DFM on peptide binder design with five objectives (hemolysis, non-fouling, solubility, half-life, binding affinity) and on enhancer DNA generation guided by class and shape. They also compare against classical MOO baselines and report ablations. Appendix D claims a formal guarantee that MOG-DFM induces positive expected improvement in the direction of the weight vector.

Significance. If the empirical results were independently confirmed, MOG-DFM would be a useful contribution to controllable discrete sequence generation. The paper includes trained base models, a planned public codebase, ablations isolating the hypercone mechanism, and some external checks with ADMET-AI and AlphaFold3/VINA. However, the formal support is definitional rather than a genuine convergence guarantee, and the main evaluations reuse the same predictors that provide guidance, so the strength of the Pareto-efficiency claims is not currently supported. The framework itself is plausible and the empirical trends are internally consistent, but the manuscript needs substantial revision of both the theoretical claim and the evaluation design before the central claims can be accepted.

major comments (3)
  1. [Appendix D and Algorithm 1] The formal guarantee is not established for the implemented algorithm. The proof takes expectation under the softmax distribution mu_i over feasible transitions, but Section 3.4 and Eq. (23) select y_best = argmax over accepted candidates and use the guided rate only to decide whether to move; these are different stochastic processes. In addition, the key inequality D(y_i,x;omega)>0 is claimed to follow from alpha_i <= Phi with Phi in (0,pi), but arccos of a dot product is nonnegative only when Phi <= pi/2; for Phi > pi/2 the cone contains vectors with negative dot product. The fallback rule in Eq. (18) selects among candidates with alpha_i < pi, which can include transitions with D<0, so the proof's assertion that the fallback chooses D>0 'by construction' is false. At best, the argument shows positive expected scalarized improvement for a softmax sampler restricted to an acute cone, which is a direct consequence of the filter rather than an independent convergence proof. It does not establish monotonic progress toward the Pareto front, since moves with positive dot product along one weight vector can degrade individual objectives.
  2. [Section 4.3 and Section B] The headline improvements are evaluated with the same predictors used for guidance. The hemolysis, non-fouling, solubility, half-life, and affinity scores in Tables 2, 3, 7, 8, 9, and 10 are produced by the models described in Section B, which are exactly the score functions used in Eq. (14) during sampling. The external ADMET-AI and AlphaFold3/VINA checks cover only solubility, half-life, and binding on selected structures; non-fouling and hemolysis have no independent check, and the binding check is limited to a small number of examples. The reported Pareto gains may therefore reflect predictor artifacts rather than genuine molecular improvements, and the claim that MOG-DFM 'consistently produces the most favorable trade-offs' is not supported by the experiments as designed.
  3. [Tables 3, 4, 7, 10] Most quantitative comparisons lack error bars, standard deviations, or significance tests. In particular, Table 4 evaluates each DNA guidance setting on only five sequences, and Table 10 varies one hyperparameter at a time on different targets, so cross-setting comparisons are confounded. Without variance estimates, it is impossible to judge whether the large half-life differences in Table 3 (for example, 27.227 h vs 4.449 h for target 1B8Q) or the ablation differences in Table 7 are meaningful or within sampling noise. The paper should report distributions or confidence intervals and use a consistent evaluation setup when comparing settings.
minor comments (5)
  1. [Table 4] The column layout is confusing: the 'Shape' label appears multiple times and the Task 1 and Task 2 values are presented side by side without a clear separator, making the ablation rows hard to read.
  2. [Section 4.1] The sentence 'the best EnhancerDFM model is achieved within 20 training epochs, while the best EnhancerDFM is obtained only in around 1400 training epochs' presumably should refer to the best Dirichlet FM model in the second clause; as written it is self-contradictory.
  3. [Eq. (16)] The angle definition in Eq. (16) is undefined when the improvement vector Delta s(y_i,x) is zero; the paper should specify the convention for zero improvement vectors.
  4. [Abstract and Section 3] The abstract and Section 3 refer to steering toward 'Pareto-efficient' solutions, while the definition of the Pareto front in Section 3 is followed by the parenthetical '(not guaranteed to be Pareto optimal)'. These statements are in tension and should be reconciled in the revised text.
  5. [Table 3] MOG-DFM is substantially slower than the baselines (43-90 s vs 8-37 s per sequence); since the conclusion emphasizes scalability, the paper should discuss whether this runtime gap is acceptable for practical use.

Circularity Check

1 steps flagged · score 6.0 of 10

Appendix D's positive-expected-improvement guarantee is the hypercone filter restated by definition, and the proof does not match the argmax sampler.

  1. self definitional [Appendix D, 'Additional Proof'; hypercone filter defined in Section 3.3, Eqs. (16)-(17)]
    "Define the set of feasible transitions (those within the hypercone of angle Φ∈(0,π)) at time t as: Yi(x,ω,Φ) := { yi∈T\{xi} | arccos( ∆s(yi,x)·ω / (∥∆s(yi,x)∥·∥ω∥) ) ≤ Φ }. ... Since each yi ∈ Yi(x,ω,Φ) satisfies arccos(...) ≤ Φ < π, it follows that D(yi,x;ω)>0 for all yi∈Yi. Moreover, µ i t(yi|x,ω)>0 by construction. Therefore, each term in the sum is strictly positive, and thus: E[∆s(xnew,x)·ω]>0."

    The accepted set Yi is defined by the condition that the angle between Δs and ω is at most Φ (Eq. 16). For any acute cone (Φ<π/2) this condition is exactly D = ‖Δs‖‖ω‖cos(angle) > 0, so the proof's conclusion—positive expected improvement in direction ω—is merely the filter's acceptance rule rewritten as an expectation. It does not follow that the sampler converges to or reaches the Pareto front, since a positive drift along one trade-off direction says nothing about Pareto efficiency. The proof is also inapplicable to the submitted algorithm: it averages under a softmax measure µ_i over all accepted candidates, whereas Section 3.4 and Algorithm 1 select y_best = argmax and transition with kernel (23); and for Φ between π/2 and π the claimed implication D>0 fails.

full rationale

The paper's only formal support for the central Pareto claim is Appendix D. That proof defines the accepted transition set as exactly the moves whose improvement vector lies within angle Φ of ω (Eq. 16), so for acute cones every accepted move has positive dot product with ω by construction; summing positive terms under any positive measure gives E[D]>0 trivially. This is a definitional restatement, not a convergence or Pareto-efficiency proof, and the softmax expectation in the proof does not match the argmax selection and Euler kernel used by Algorithm 1/Eq. (23). The empirical sections are not forced: solubility and half-life are cross-checked with the independently trained ADMET-AI, binding with AlphaFold3 and AutoDock VINA, and ablations/baseline comparisons provide independent evidence that the guidance changes outputs. Self-citations (PepTune [23], Gumbel-Softmax FM [29]) are contextual and not load-bearing for the MOG-DFM mechanism. Overall, the formal Pareto guarantee is partially circular/definitional, while the method has substantial independent empirical content.

Assumptions & free parameters 12 free parameters · 6 assumptions · 0 invented entities

The central claim rests on the pretrained flow matching base models, the property predictor score functions, and a set of hand-chosen hyperparameters. The most consequential input is the reliability of the score models, because they appear on both sides of the experiment: they guide sampling and they measure the reported outcomes. The formal proof also relies on a feasibility assumption about the hypercone acceptance set, and its conclusion is largely encoded in the filter definition.

free parameters (12)
  • lambda_direction_weight = 1.0
    Balances rank-based improvement and directional alignment in Eq (14); chosen by hand in Section C.1.
  • beta_guidance_strength = 1.0
    Scales the exponential reweighting of the base velocity in Eq (15); chosen by hand.
  • initial_hypercone_angle_Phi_init = 45 degrees
    Initial angular acceptance threshold; sensitivity shown in Table 10.
  • hypercone_bounds_Phi_min_Phi_max = [15, 75] degrees
    Clipping bounds for adaptive cone angle in Eq (21).
  • target_rejection_rate_tau = 0.3
    EMA target rejection rate that controls cone expansion and contraction in Eq (21).
  • ema_smoothing_alpha_r = 0.5
    Smoothing coefficient for rejection-rate EMA in Eq (20).
  • cone_learning_rate_eta = 1.0
    Learning rate for hypercone angle update in Eq (21).
  • num_div_Das_Dennis = 64
    Number of subdivisions for weight-vector lattice; controls Pareto direction sampling in Eq (11).
  • importance_vector_peptide = [1, 1, 1, 0.5, 0.2]
    Hand-set to inversely approximate property ranges; in Section C.1.
  • importance_vector_dna = [1, 10] and [1, 100]
    Hand-set inversely to shape-value ranges for HelT and Rise tasks; Section C.2.
  • sampling_steps_T = 100 (peptide), 800 (DNA)
    Discretization steps for Euler sampling; sensitivity in Table 10.
  • mixture_path_polynomial_exponent = 2.0
    Scheduler exponent for both base models in Appendix A; model choice, not fitted to target results.
assumptions (6)
  • standard math CTMC rate conditions and mixture path parametrization for discrete flow matching (Gat et al.) are valid for the trained base models.
    Section 2 builds the guided sampler on these rate conditions and the mixture posterior in Eq (7).
  • standard math The Euler-Maruyama analogue in Eq (23) approximates the continuous-time Markov chain at step size h=1/T.
    Used in Algorithm 1 Step 3; standard numerical treatment of CTMCs.
  • domain assumption The scalar score functions s_n are accurate enough proxies for the biological properties being optimized.
    All guidance and most evaluation rely on these predictors; validation F1 values are 0.58 to 0.71 and Spearman values 0.64 to 0.86 (Section B), and only solubility and half-life are externally cross-checked with ADMET-AI.
  • domain assumption The pretrained base model assigns probability mass to sequences from which Pareto-relevant sequences can be reached by single-token substitutions.
    MOG-DFM updates one position per iteration and only reweights existing token candidates; if the base distribution cannot reach good regions through such moves, the sampler cannot find them. Algorithm 1 Step 1.
  • ad hoc to paper In Section D the set of feasible hypercone transitions Yi is non-empty and contains moves with D(y_i,x;omega)>0.
    The proof assumes this; when all moves have negative or zero directional improvement the algorithm self-transitions (Section 3.3), so the positive-expected-improvement argument does not cover the actual failure mode.
  • ad hoc to paper Accepted transitions within angle Phi<pi have strictly positive directional improvement, which is the source of the proof's main inequality.
    This makes the Section D claim true by construction rather than by an independent convergence argument; it is a property of the filter, not a theorem about Pareto optimality.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Multi-Objective-Guided Discrete Flow Matching for Controllable Biological Sequence Design." pith.science (2026). https://pith.science/paper/VVERKBFB

@misc{pith2026250507086,
  author       = {Pith},
  title        = {Pith review of: Multi-Objective-Guided Discrete Flow Matching for Controllable Biological Sequence Design},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/VVERKBFB}},
  note         = {Machine review of arXiv:2505.07086}
}
read the original abstract

Designing biological sequences that satisfy multiple, often conflicting, functional and biophysical criteria remains a central challenge in biomolecule engineering. While discrete flow matching models have recently shown promise for efficient sampling in high-dimensional sequence spaces, existing approaches address only single objectives or require continuous embeddings that can distort discrete distributions. We present Multi-Objective-Guided Discrete Flow Matching (MOG-DFM), a general framework to steer any pretrained discrete flow matching generator toward Pareto-efficient trade-offs across multiple scalar objectives. At each sampling step, MOG-DFM computes a hybrid rank-directional score for candidate transitions and applies an adaptive hypercone filter to enforce consistent multi-objective progression. We also trained two unconditional discrete flow matching models, PepDFM for diverse peptide generation and EnhancerDFM for functional enhancer DNA generation, as base generation models for MOG-DFM. We demonstrate MOG-DFM's effectiveness in generating peptide binders optimized across five properties (hemolysis, non-fouling, solubility, half-life, and binding affinity), and in designing DNA sequences with specific enhancer classes and DNA shapes. In total, MOG-DFM proves to be a powerful tool for multi-property-guided biomolecule sequence design.

Figures

Figures reproduced from arXiv: 2505.07086 by the authors.

Figure 1
Figure 1. Visualization for MOG-DFM algorithm. 3 Multi-Objective Guided Discrete Flow Matching MOG-DFM (Multi-Objective Guided Discrete Flow Matching) operates under the same setting as discrete flow matching described in the previous section. Suppose we have a pre-trained discrete flow matching model that defines a CTMC with a factorized velocity field u i t (y i , x) , which transports probability mass from an initial distr… view at source ↗
Figure 2
Figure 2. (A), (B) Complex structures of PDB 5AZ8 with a MOG-DFM-designed binder and its pre-existing binder. (C), (D) Complex structures of two target proteins without pre-existing binders (OX1R, EWS::FLI1) with MOG-DFM-designed binders. Five property scores are shown for each binder, along with the ipTM score from AlphaFold3 and docking score from AutoDock VINA. Interacting residues on the target are visualized. (E) Plots s… view at source ↗
Figure 3
Figure 3. (A) The Hamming distance of sampled peptides of different lengths to the peptides of the same length in the test set. (B) The Shannon Entropy of sampled peptides of different lengths to the peptides of the same length in the test set. 24 [PITH_FULL_IMAGE:figures/full_fig_p024_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Complex structures of target proteins with pre-existing binders. (A)-(B) 1B8Q, (C)-(D) 1E6I, (E)-(F) 3IDJ, (G)-(H) 7JVS. Each panel shows the complex structure of the target with either a MOG-DFM￾designed binder or its pre-existing binder. For each binder, five propert…
Figure 5
Figure 5. Figure 5: Complex structures of target proteins without pre-existing binders. (A)-(C) AMHR2, (D)-(E) EWS::FLI1, (F) MYC, (G) DUSP12. Each panel shows the complex structure of the target with a MOG-DFM￾designed binder. For each binder, five property scores are provided, as well a…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

53 extracted references · 44 canonical work pages

  1. [1]

    Application of combinatorial optimization strategies in synthetic biology

    Gita Naseri and Mattheos AG Koffas. Application of combinatorial optimization strategies in synthetic biology. Nature communications, 11(1):2446, 2020

    work page 2020

  2. [2]

    Designing strong inducible synthetic promoters in yeasts

    Masahiro Tominaga, Yoko Shima, Kenta Nozaki, Yoichiro Ito, Masataka Someda, Yuji Shoya, Noritaka Hashii, Chihiro Obata, Miho Matsumoto-Kitano, Kohei Suematsu, et al. Designing strong inducible synthetic promoters in yeasts. Nature Communications, 15(1):10653, 2024

    work page 2024

  3. [3]

    Crispr guide rna design for research applications

    Stephanie E Mohr, Yanhui Hu, Benjamin Ewen-Campen, Benjamin E Housden, Raghuvir Viswanatha, and Norbert Perrimon. Crispr guide rna design for research applications. The FEBS journal, 283(17):3232–3238, 2016

    work page 2016

  4. [4]

    Genome-wide crispr guide rna design and specificity analysis with guidescan2

    Henri Schmidt, Minsi Zhang, Dimitar Chakarov, Vineet Bansal, Haralambos Mourelatos, Francisco J Sánchez-Rivera, Scott W Lowe, Andrea Ventura, Christina S Leslie, and Yuri Pritykin. Genome-wide crispr guide rna design and specificity analysis with guidescan2. Genome biology, 26(1):1–25, 2025

    work page 2025

  5. [5]

    Synthetic promoters in gene therapy: Design approaches, features and applications

    Valentin Artemyev, Anna Gubaeva, Anastasiia Iu Paremskaia, Amina A Dzhioeva, Andrei Deviatkin, Sofya G Feoktistova, Olga Mityaeva, and Pavel Yu V olchkov. Synthetic promoters in gene therapy: Design approaches, features and applications. Cells, 13(23):1963, 2024

    work page 1963

  6. [6]

    All-small-molecule organic solar cells with over 14% efficiency by optimizing hierarchical morphologies

    Ruimin Zhou, Zhaoyan Jiang, Chen Yang, Jianwei Yu, Jirui Feng, Muhammad Abdullah Adil, Dan Deng, Wenjun Zou, Jianqi Zhang, Kun Lu, et al. All-small-molecule organic solar cells with over 14% efficiency by optimizing hierarchical morphologies. Nature communications, 10 (1):5393, 2019

    work page 2019

  7. [7]

    Novel strategies to optimize the amplification of single-stranded dna

    Atef Nehdi, Nosaibah Samman, Vanessa Aguilar-Sánchez, Azer Farah, Emre Yurdusev, Mo- hamed Boudjelal, and Jonathan Perreault. Novel strategies to optimize the amplification of single-stranded dna. Frontiers in Bioengineering and Biotechnology, 8:401, 2020

    work page 2020

  8. [8]

    Therapeutic strategies to reduce the toxicity of misfolded protein oligomers

    Ryan P Kreiser, Aidan K Wright, Natalie R Block, Jared E Hollows, Lam T Nguyen, Kathleen LeForte, Benedetta Mannini, Michele Vendruscolo, and Ryan Limbocker. Therapeutic strategies to reduce the toxicity of misfolded protein oligomers. International journal of molecular sciences, 21(22):8651, 2020

    work page 2020

Show all 53 references
  1. [9]

    Toxinpred2: an improved method for predicting toxicity of proteins

    Neelam Sharma, Leimarembi Devi Naorem, Shipra Jain, and Gajendra PS Raghava. Toxinpred2: an improved method for predicting toxicity of proteins. Briefings in bioinformatics , 23(5): bbac174, 2022

    2022

  2. [10]

    Neural network conditioned to produce thermophilic protein sequences can increase thermal stability

    Evan Komp, Christian Phillips, Lauren M Lee, Shayna M Fallin, Humood N Alanzi, Marlo Zorman, Michelle E McCully, and David AC Beck. Neural network conditioned to produce thermophilic protein sequences can increase thermal stability. Scientific Reports, 15(1):14124, 2025

    2025

  3. [11]

    The toxicity of protein aggregates: new insights into the mechanisms, 2023

    Alessandra Bigi, Eva Lombardo, Roberta Cascella, and Cristina Cecchi. The toxicity of protein aggregates: new insights into the mechanisms, 2023

    2023

  4. [12]

    Misfolded protein oligomers: Mechanisms of formation, cytotoxic effects, and pharmacological approaches against protein misfolding diseases

    Dillon J Rinauro, Fabrizio Chiti, Michele Vendruscolo, and Ryan Limbocker. Misfolded protein oligomers: Mechanisms of formation, cytotoxic effects, and pharmacological approaches against protein misfolding diseases. Molecular Neurodegeneration, 19(1):20, 2024

    2024

  5. [13]

    Multiobjective optimization using evolutionary algorithms—a comparative case study

    Eckart Zitzler and Lothar Thiele. Multiobjective optimization using evolutionary algorithms—a comparative case study. In International conference on parallel problem solving from nature, pages 292–301. Springer, 1998. 12

    1998

  6. [14]

    Multi-objective optimisation using evolutionary algorithms: an introduction

    Kalyanmoy Deb. Multi-objective optimisation using evolutionary algorithms: an introduction. In Multi-objective evolutionary optimisation for product design and manufacturing, pages 3–34. Springer, 2011

    2011

  7. [15]

    Combo: An efficient bayesian optimization library for materials science

    Tsuyoshi Ueno, Trevor David Rhone, Zhufeng Hou, Teruyasu Mizoguchi, and Koji Tsuda. Combo: An efficient bayesian optimization library for materials science. Materials discovery, 4:18–21, 2016

    2016

  8. [16]

    Bayesian optimization with evolutionary and structure-based regularization for directed protein evolution

    Trevor S Frisby and Christopher James Langmead. Bayesian optimization with evolutionary and structure-based regularization for directed protein evolution. Algorithms for Molecular Biology, 16(1):13, 2021

    2021

  9. [17]

    Multi-objective de novo drug design with conditional graph generative model

    Yibo Li, Liangren Zhang, and Zhenming Liu. Multi-objective de novo drug design with conditional graph generative model. Journal of cheminformatics, 10:1–24, 2018

    2018

  10. [18]

    Combining multi-objective evolutionary algorithms with deep generative models towards focused molecular design

    Tiago Sousa, João Correia, Vitor Pereira, and Miguel Rocha. Combining multi-objective evolutionary algorithms with deep generative models towards focused molecular design. In Applications of Evolutionary Computation: 24th International Conference, EvoApplications 2021, Held as...

    2021

  11. [19]

    Proud: Pareto-guided diffusion model for multi-objective generation

    Yinghua Yao, Yuangang Pan, Jing Li, Ivor Tsang, and Xin Yao. Proud: Pareto-guided diffusion model for multi-objective generation. Machine Learning, 113(9):6511–6538, 2024

    2024

  12. [20]

    Paretoflow: Guided flows in multi-objective optimization

    Ye Yuan, Can Chen, Christopher Pal, and Xue Liu. Paretoflow: Guided flows in multi-objective optimization. arXiv preprint arXiv:2412.03718, 2024

    2024 arXiv

  13. [21]

    Challenges of continuous global optimization in molecular structure prediction

    Gleb Beliakov and Kieran F Lim. Challenges of continuous global optimization in molecular structure prediction. European journal of operational research, 181(3):1198–1213, 2007

    2007

  14. [22]

    A continuous relaxation for discrete bayesian optimiza- tion

    Richard Michael, Simon Bartels, Miguel González-Duque, Yevgen Zainchkovskyy, Jes Frellsen, Søren Hauberg, and Wouter Boomsma. A continuous relaxation for discrete bayesian optimiza- tion. arXiv preprint arXiv:2404.17452, 2024

    2024 arXiv

  15. [23]

    Peptune: De novo generation of therapeutic peptides with multi-objective-guided discrete diffusion

    Sophia Tang, Yinuo Zhang, and Pranam Chatterjee. Peptune: De novo generation of therapeutic peptides with multi-objective-guided discrete diffusion. Proceedings of the 42nd International Conference on Machine Learning (ICML), 2025

    2025

  16. [24]

    Simple and effective masked diffusion language models

    Subham Sekhar Sahoo, Marianne Arriola, Yair Schiff, Aaron Gokaslan, Edgar Marroquin, Justin T Chiu, Alexander Rush, and V olodymyr Kuleshov. Simple and effective masked diffusion language models. Advances in Neural Information Processing Systems, 2024

    2024

  17. [25]

    Discrete flow matching

    Itai Gat, Tal Remez, Neta Shaul, Felix Kreuk, Ricky TQ Chen, Gabriel Synnaeve, Yossi Adi, and Yaron Lipman. Discrete flow matching. Advances in Neural Information Processing Systems, 37:133345–133385, 2024

    2024

  18. [26]

    Exploring discrete flow matching for 3d de novo molecule generation

    Ian Dunn and David Ryan Koes. Exploring discrete flow matching for 3d de novo molecule generation. ArXiv, pages arXiv–2411, 2024

    2024

  19. [27]

    Dirichlet flow matching with applications to dna sequence design.Proceedings of the 41st International Conference on Machine Learning (ICML), 2024

    Hannes Stark, Bowen Jing, Chenyu Wang, Gabriele Corso, Bonnie Berger, Regina Barzilay, and Tommi Jaakkola. Dirichlet flow matching with applications to dna sequence design.Proceedings of the 41st International Conference on Machine Learning (ICML), 2024

    2024

  20. [28]

    Fisher flow matching for generative modeling over discrete data

    Oscar Davis, Samuel Kessler, Mircea Petrache, Ismail Ceylan, Michael Bronstein, and Joey Bose. Fisher flow matching for generative modeling over discrete data. Advances in Neural Information Processing Systems, 37:139054–139084, 2024

    2024

  21. [29]

    Gumbel-softmax flow matching with straight-through guidance for controllable biological sequence generation

    Sophia Tang, Yinuo Zhang, Alexander Tong, and Pranam Chatterjee. Gumbel-softmax flow matching with straight-through guidance for controllable biological sequence generation. arXiv preprint arXiv:2503.17361, 2025

    2025 arXiv

  22. [30]

    Unlocking guid- ance for discrete state-space diffusion and flow models

    Hunter Nisonoff, Junhao Xiong, Stephan Allenspach, and Jennifer Listgarten. Unlocking guid- ance for discrete state-space diffusion and flow models. Proceedings of the 13th International Conference on Learning Representations (ICLR), 2025. 13

    2025

  23. [31]

    Pepnn: a deep attention model for the identification of peptide binding sites

    Osama Abdin, Satra Nim, Han Wen, and Philip M Kim. Pepnn: a deep attention model for the identification of peptide binding sites. Communications biology, 5(1):503, 2022

    2022

  24. [32]

    Biolip2: an updated structure database for biologically relevant ligand–protein interactions

    Chengxin Zhang, Xi Zhang, Peter L Freddolino, and Yang Zhang. Biolip2: an updated structure database for biologically relevant ligand–protein interactions. Nucleic Acids Research, 52(D1): D404–D412, 2024

    2024

  25. [33]

    Learning to de- sign protein-protein interactions with enhanced generalization.arXiv preprint arXiv:2310.18515, 2023

    Anton Bushuiev, Roman Bushuiev, Petr Kouba, Anatolii Filkin, Marketa Gabrielova, Michal Gabriel, Jiri Sedlar, Tomas Pluskal, Jiri Damborsky, Stanislav Mazurenko, et al. Learning to de- sign protein-protein interactions with enhanced generalization.arXiv preprint arXiv:2310.18515, 2023

    arXiv

  26. [34]

    Accurate structure prediction of biomolecular interactions with alphafold 3

    Josh Abramson, Jonas Adler, Jack Dunger, Richard Evans, Tim Green, Alexander Pritzel, Olaf Ronneberger, Lindsay Willmore, Andrew J Ballard, Joshua Bambrick, et al. Accurate structure prediction of biomolecular interactions with alphafold 3. Nature, 630(8016):493–500, 2024

    2024

  27. [35]

    Oleg Trott and Arthur J Olson. Autodock vina: improving the speed and accuracy of docking with a new scoring function, efficient optimization, and multithreading.Journal of computational chemistry, 31(2):455–461, 2010

    2010

  28. [36]

    Admet-ai: a machine learning admet platform for evaluation of large-scale chemical libraries

    Kyle Swanson, Parker Walther, Jeremy Leitz, Souhrid Mukherjee, Joseph C Wu, Rabindra V Shivnaraine, and James Zou. Admet-ai: a machine learning admet platform for evaluation of large-scale chemical libraries. Bioinformatics, 40(7):btae416, 2024

    2024

  29. [37]

    An evolutionary many-objective optimization algorithm using reference-point-based nondominated sorting approach, part i: solving problems with box constraints

    Kalyanmoy Deb and Himanshu Jain. An evolutionary many-objective optimization algorithm using reference-point-based nondominated sorting approach, part i: solving problems with box constraints. IEEE transactions on evolutionary computation, 18(4):577–601, 2013

    2013

  30. [38]

    Sms-emoa: Multiobjective selection based on dominated hypervolume.European journal of operational research, 181(3):1653–1669, 2007

    Nicola Beume, Boris Naujoks, and Michael Emmerich. Sms-emoa: Multiobjective selection based on dominated hypervolume.European journal of operational research, 181(3):1653–1669, 2007

    2007

  31. [39]

    Spea2: Improving the strength pareto evolutionary algorithm

    Eckart Zitzler, Marco Laumanns, and Lothar Thiele. Spea2: Improving the strength pareto evolutionary algorithm. TIK report, 103, 2001

    2001

  32. [40]

    Mopso: A proposal for multiple objective particle swarm optimization

    CA Coello Coello and Maximino Salazar Lechuga. Mopso: A proposal for multiple objective particle swarm optimization. In Proceedings of the 2002 Congress on Evolutionary Computation. CEC’02 (Cat. No. 02TH8600), volume 2, pages 1051–1056. IEEE, 2002

    2002

  33. [41]

    Predicting dna structure using a deep learning method

    Jinsen Li, Tsu-Pei Chiu, and Remo Rohs. Predicting dna structure using a deep learning method. Nature communications, 15(1):1243, 2024

    2024

  34. [42]

    U-net: Convolutional networks for biomedical image segmentation

    Olaf Ronneberger, Philipp Fischer, and Thomas Brox. U-net: Convolutional networks for biomedical image segmentation. In Medical image computing and computer-assisted intervention–MICCAI 2015: 18th international conference, Munich, Germany, October 5-9, 2015, proceedings, part ...

    2015

  35. [43]

    Evolutionary-scale prediction of atomic-level protein structure with a language model

    Zeming Lin, Halil Akin, Roshan Rao, Brian Hie, Zhongkai Zhu, Wenting Lu, Nikita Smetanin, Robert Verkuil, Ori Kabeli, Yaniv Shmueli, et al. Evolutionary-scale prediction of atomic-level protein structure with a language model. Science, 379(6637):1123–1130, 2023

    2023

  36. [44]

    Interpretation of allele-specific chromatin accessibility us- ing cell state–aware deep learning

    Zeynep Kalender Atak, Ibrahim Ihsan Taskiran, Jonas Demeulemeester, Christopher Flerin, David Mauduit, Liesbeth Minnoye, Gert Hulselmans, Valerie Christiaens, Ghanem-Elias Ghanem, Jasper Wouters, et al. Interpretation of allele-specific chromatin accessibility us- ing cell sta...

    2021

  37. [45]

    Transposition of native chromatin for fast and sensitive epigenomic profiling of open chromatin, dna-binding proteins and nucleosome position

    Jason D Buenrostro, Paul G Giresi, Lisa C Zaba, Howard Y Chang, and William J Greenleaf. Transposition of native chromatin for fast and sensitive epigenomic profiling of open chromatin, dna-binding proteins and nucleosome position. Nature methods, 10(12):1213–1218, 2013

    2013

  38. [46]

    Pepland: a large-scale pre-trained peptide representation model for a comprehensive landscape of both canonical and non-canonical amino acids

    Ruochi Zhang, Haoran Wu, Yuting Xiu, Kewei Li, Ningning Chen, Yu Wang, Yan Wang, Xin Gao, and Fengfeng Zhou. Pepland: a large-scale pre-trained peptide representation model for a comprehensive landscape of both canonical and non-canonical amino acids. arXiv preprint arXiv:2311...

    2023

  39. [47]

    Peptidebert: A language model based on transformers for peptide property prediction

    Chakradhar Guntuboina, Adrita Das, Parisa Mollaei, Seongwon Kim, and Amir Barati Farimani. Peptidebert: A language model based on transformers for peptide property prediction. The Journal of Physical Chemistry Letters, 14(46):10427–10434, 2023

    2023

  40. [48]

    Pedregosa, G

    F. Pedregosa, G. Varoquaux, A. Gramfort, V . Michel, B. Thirion, O. Grisel, M. Blondel, P. Prettenhofer, R. Weiss, V . Dubourg, J. Vanderplas, A. Passos, D. Cournapeau, M. Brucher, M. Perrot, and E. Duchesnay. Scikit-learn: Machine learning in Python. Journal of Machine Learni...

    2011

  41. [49]

    Optuna: A next-generation hyperparameter optimization framework

    Takuya Akiba, Shotaro Sano, Toshihiko Yanase, Takeru Ohta, and Masanori Koyama. Optuna: A next-generation hyperparameter optimization framework. In International Conference on Knowledge Discovery and Data Mining, pages 2623–2631, 2019

    2019

  42. [50]

    Peplife: a repository of the half-life of peptides

    Deepika Mathur, Satya Prakash, Priya Anand, Harpreet Kaur, Piyush Agrawal, Ayesha Mehta, Rajesh Kumar, Sandeep Singh, and Gajendra PS Raghava. Peplife: a repository of the half-life of peptides. Scientific reports, 6(1):36617, 2016

    2016

  43. [51]

    Peptherdia: database and structural composition analysis of approved peptide therapeutics and diagnostics

    Vera D’Aloisio, Paolo Dognini, Gillian A Hutcheon, and Christopher R Coxon. Peptherdia: database and structural composition analysis of approved peptide therapeutics and diagnostics. Drug Discovery Today, 26(6):1409–1419, 2021

    2021

  44. [52]

    Thpdb2: compilation of fda approved therapeutic peptides and proteins

    Shipra Jain, Srijanee Gupta, Sumeet Patiyal, and Gajendra PS Raghava. Thpdb2: compilation of fda approved therapeutic peptides and proteins. Drug Discovery Today, page 104047, 2024

    2024

  45. [53]

    Mega-scale experimental analysis of protein folding stability in biology and design

    Kotaro Tsuboyama, Justas Dauparas, Jonathan Chen, Elodie Laine, Yasser Mohseni Behbahani, Jonathan J Weinstein, Niall M Mangan, Sergey Ovchinnikov, and Gabriel J Rocklin. Mega-scale experimental analysis of protein folding stability in biology and design. Nature, 620(7973): 43...

    2023

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.