Leveraging generative hallucination and biophysics-informed modeling for unified biomolecular sequence-structure co-design

Xuefeng Liu; Mingxuan Cao; Xiao Luo; Songhao Jiang; Tobin Sosnick; Jinbo Xu; Louis Maher; Rick Stevens

q-bio.QMAug 18, 2026

Leveraging generative hallucination and biophysics-informed modeling for unified biomolecular sequence-structure co-design

Authors: Xuefeng Liu, Mingxuan Cao, Xiao Luo, Songhao Jiang, Tobin Sosnick, Jinbo Xu, Louis Maher, Rick Stevens

Organizations: Department of Medicine, University of Florida · Department of Computer Science, University of Chicago · Data Science Institute, University of Chicago · Department of Pathology, University of Chicago · Department of Biochemistry and Molecular Biology, University of Chicago · Toyota Technological Institute at Chicago · Department of Biochemistry and Molecular Biology, Mayo Clinic · Argonne National Laboratory

Abstract

Biomolecular design underpins applications from molecular recognition to therapeutics and synthetic biology, yet de novo interaction design remains challenging-especially for DNA/RNA, underexplored non-protein modalities with scarce, heterogeneous complex data and sharper geometric and chemical constraints. We introduce MCTH (Monte Carlo Tree Hallucination), an inference-only framework that casts all-atom sequence-structure co-design as uncertainty-aware planning over hallucinated states from pretrained folding and inverse-folding models, with optional biophysical control within the same decision loop. MCTH treats these models as frozen black-box operators and uses Monte Carlo Tree Search to allocate a fixed inference budget across competing design trajectories, incorporating model confidence and uncertainty, as well as cross-expert consensus/disagreement when multiple predictors are available. Across protein-RNA, protein-DNA, protein-protein, and protein-ligand design, matched-budget experiments show that adaptive search improves over simpler sampling and cycling strategies, while held-out AlphaFold3 and Chai-1 evaluations demonstrate transfer beyond the search-time oracle. MCTH provides a shared planning layer across modalities while allowing task-specific folding, inverse-folding, and biophysical modules, requiring no fine-tuning or backpropagation through component models.

Figures & tables

Figure 1: The MCTH framework: search-guided sequence–structure co-design. Left: UCT-style Monte Carlo Tree Search performs global planning over states $s=(S,X)$ via selection and backpropagation. Right: each expansion instantiates a hallucination–evaluation cycle that proposes new sequences (inverse folding), generates corresponding structures (diffusion-based hallucination), and evaluates candidates using model confidence signals (e.g., ipTM/pLDDT), which are then backed up to guide subsequent search. Confidence-guided masking is an optional refinement operator and is disabled in the minimal plug-and-play configuration used in this work.

Table 1: Protein–RNA aptamer design. Left: matched 100-candidate evaluation under the Boltz-2 search-time oracle. Right: independent AlphaFold3 re-scoring of final designs. Values are ipTM; higher is better.

Table 2: Protein–DNA aptamer design. Left: matched 100-candidate evaluation under the Boltz-2 search-time oracle; the two 7XVN chain-B cases are excluded because Boltz-2 collapses for all compared designs. Right: independent AlphaFold3 re-scoring on all 10 targets, under which the two chain-B cases are non-degenerate. Values are ipTM; higher is better.

Target	BoltzDesign1@100	ODesign@100	Complexa@100	MCTH @100
1DJS	0.758	0.944	0.907	0.933
1MOX	0.291	0.901	0.845	0.929
1XIW	0.884	0.940	0.946	0.946
2GY7	0.189	0.672	0.930	0.862
2IFG	0.527	0.730	0.356	0.806
3DI3	0.508	0.901	0.926	0.947

Table 3: Matched-budget protein-binder optimization on CaoData. All methods use up to 100 folding/evaluator calls per target. ODesign and Complexa candidates are evaluated under the common Boltz-2 refolding protocol, while BoltzDesign1 uses its native Boltz evaluation during optimization. † BoltzDesign1 mean is over 11 targets because 5U8R is unavailable. Values are best ipTM per target; higher is better.

Method	Outputs/target	Passing/valid	Pass rate	Targets $\geq$ 1 pass
Dual-expert MCTH	10	41/110	37.3%	8/11
RFDiffusion	10	22/110	20.0%	8/11

Table 5: Matched final-design yield under the same held-out binder-quality filter.

Appendix figures & tables15 assets

Supplementary material from the paper’s appendix.

Appendix

Method family / example	Plans over seq.–struct. states	Multi-modality	Biophysics in-loop	Fine-tuning-free
Rosetta / MD-style optimization	$\times$	Limited	✓	✓
RFdiffusion / ODesign	$\times$	$\triangle$	$\times$	✓
ProteinMPNN / inverse folding	$\times$	Limited	$\times$	✓
BindCraft / BoltzDesign	$\triangle$	Limited	$\triangle$	✓
MCTD-ME	✓	Protein only	$\times$	✓
MCTH	✓	✓	✓	✓

Appendix

Table 6: High-level comparison with related method families. A check mark indicates an explicit/core component,

\triangle

indicates partial or task-specific support, and

\times

indicates absence.

Component	Parameter	Value
Tree search	Fold-call budget $T$	100
	Expansion width $K$	4
	Maximum tree depth $D_{\max}$	5
	Maximum MCTS iterations	50
	Fold calls per iteration	8
	Early-stop patience	10 iterations without improvement

Appendix

Table 8: Main hyperparameters and configuration.

Variant	7WKP A $\rightarrow$ B	7YEW A $\rightarrow$ B	8GXB C $\rightarrow$ B	Mean	$\Delta$ vs. Full
MCTH (full)	0.902 $\pm$ 0.010	0.979 $\pm$ 0.002	0.858 $\pm$ 0.006	0.913	0.000
Exploration-off $(c_{p}=0)$	0.756 $\pm$ 0.007	0.972 $\pm$ 0.002	0.791 $\pm$ 0.040	0.839	-0.073
Greedy cycling	0.750 $\pm$ 0.004	0.957 $\pm$ 0.003	0.646 $\pm$ 0.012	0.784	-0.129
Best-of- $N$ random	0.752 $\pm$ 0.030	0.953 $\pm$ 0.001	0.629 $\pm$ 0.010	0.778	-0.135

Appendix

Table 12: Matched-budget ablation on RNA aptamer design. All values are mean

\pm

standard deviation over three seeds.

Variant	7XVN A_N	7YUK A_C	8PMF A_B	Mean	$\Delta$ vs. Full
MCTH (full)	0.955 $\pm$ 0.002	0.946 $\pm$ 0.008	0.964 $\pm$ 0.005	0.955	0.000
Exploration-off $(c_{p}=0)$	0.909 $\pm$ 0.011	0.904 $\pm$ 0.009	0.930 $\pm$ 0.009	0.914	-0.041
Greedy cycling	0.756 $\pm$ 0.014	0.909 $\pm$ 0.018	0.819 $\pm$ 0.006	0.828	-0.127
Best-of- $N$ random	0.753 $\pm$ 0.010	0.900 $\pm$ 0.006	0.834 $\pm$ 0.005	0.829	-0.126

Appendix

Table 13: Matched-budget ablation on DNA aptamer design. All values are mean

\pm

standard deviation over three seeds.

Variant	3DI3	4O3V	Mean	$\Delta$ vs. Full
MCTH (full)	0.912 $\pm$ 0.002	0.964 $\pm$ 0.002	0.938	0.000
Exploration-off $(c_{p}=0)$	0.897 $\pm$ 0.008	0.954 $\pm$ 0.016	0.925	-0.013
Greedy cycling	0.901 $\pm$ 0.023	0.955 $\pm$ 0.013	0.928	-0.011
Best-of- $N$ random	0.803 $\pm$ 0.005	0.900 $\pm$ 0.008	0.851	-0.087

Appendix

Table 14: Matched-budget ablation on protein–protein binder design. All values are mean

\pm

standard deviation over three seeds.

Variant	Overall mean	Protein mean	RNA mean	DNA mean
MCTH (full)	0.935	0.938	0.913	0.955
Exploration-off $(c_{p}=0)$	0.889	0.925	0.839	0.914
Greedy cycling	0.836	0.928	0.784	0.828
Best-of- $N$ random	0.815	0.851	0.778	0.829

Appendix

Table 15: Summary of matched-budget ablations across all targets.

Variant	Final mean	Never improve	Avg. last improvement
MCTH (full)	0.935	0.0%	46.6
Exploration-off ( $c_{p}=0$ )	0.889	61.9%	29.3
Greedy cycling	0.836	53.8%	23.5
Best-of- $N$	0.815	44.4%	33.3

Appendix

Table 16: Effective convergence at

\epsilon=0.002

under the matched folding-call budget. “Never improve” denotes runs with no improvement above

\epsilon

after initialization.

Modality	$n$ designs	Within-target $\rho$	95% CI
DNA aptamer	50	0.487	$[0.29,\;0.67]$
Protein binder	255	0.252	$[0.05,\;0.46]$
RNA aptamer	77	0.173	$[-0.06,\;0.36]$

Appendix

Table 17: Design-level agreement between Boltz-2 and AlphaFold3 interface scores. Correlations are computed within target and confidence intervals are obtained by bootstrap over targets.

Modality	Top-1 hit	Random	Prec.@25%	Random
DNA aptamer	0.600	0.200	0.600	0.200
Protein binder	0.167	0.053	0.427	0.241
RNA aptamer	0.286	0.091	0.143	0.273

Appendix

Table 18: Screening utility of Boltz-2 for selecting designs favored by the held-out predictor. Random columns give the corresponding chance-level expectation under the same target-level candidate-set sizes.

Archive	Passing/valid	Pass rate	Targets $\geq 1$ pass
Dual-expert MCTH, top-5	27/55	49.1%	8/11
Dual-expert MCTH, top-10	41/110	37.3%	8/11
Dual-expert MCTH, top-25	84/275	30.5%	9/11

Appendix

Table 19: Final-archive yield for dual-expert MCTH. Candidates are ranked only by the in-loop joint reward before applying the held-out filter.

Method	Outputs/target	Passing/valid	Pass rate	Targets $\geq 1$ pass
Dual-expert MCTH	10	41/110	37.3%	8/11
RFDiffusion	10	22/110	20.0%	8/11

Appendix

Table 20: Matched final-design yield using ten outputs per target and the same held-out binder-quality filter.

Target	Chai pass/10	Chai mean/best ipTM	AF2 mean/best ipTM
1DJS	1/10	0.327 / 0.594	0.166 / 0.278
1MOX	0/10	0.252 / 0.498	0.148 / 0.240
1XIW	6/10	0.540 / 0.808	0.536 / 0.840
2GY7	5/10	0.519 / 0.729	0.419 / 0.669
2IFG	2/10	0.288 / 0.700	0.259 / 0.783
3DI3	0/10	0.228 / 0.258	0.233 / 0.616

Appendix

Table 21: Per-target RFDiffusion held-out evaluation for the matched top-10 comparison. Mean/best values are computed over the ten outputs for each target.

Method	$n$	$\Delta G_{\mathrm{sep}}$	$\Delta G/100\,\text{\AA}^{2}$	Packstat	$\Delta$ unsat H-bonds
MCTH	6	-71.2	-3.12	0.577	5.7
BindCraft	5	-65.7	-3.40	0.600	5.2
RFDiffusion	6	-54.2	-3.13	0.602	6.5

Appendix

Table 22: Oracle-independent Rosetta InterfaceAnalyzer evaluation of protein-binder designs. More negative interface energy and higher packstat are favorable; fewer buried unsatisfied hydrogen bonds are preferred.

Modality	Targets	Mean folding calls	Mean wall-clock
Protein binder	12	98.1	$\sim$ 9.0 h
RNA aptamer	7	52.4	$\sim$ 0.8 h
DNA aptamer	8	66.1	$\sim$ 1.1 h

Appendix

Table 23: Realized compute for the main MCTH experiments. DNA reports the eight targets used in the non-degenerate Boltz-2 comparison. Wall-clock values are means from the corresponding run logs.

Fold call	1	5	10	25	50	100
Protein binder	0.614	0.724	0.768	0.803	0.830	0.864

Appendix

Table 24: Protein-binder performance as a function of folding-call budget. Values are mean held-forward best ipTM across targets.

Explore similar work

Chamaileon: Cross-Context Binder Design with Contextualized Modeling and Mixed Sampling
Jul 26, 2026Hengyuan Cao, Shizhuo Cheng, Mingxuan Liu +5Molecular Generation Protein Design
Zhejiang University, Hangzhou, China · Fudan University, Shanghai, China · Shanghai Institute for Advanced Study Zhejiang University, Shanghai, China +1
SimpleDesign: A Joint Model for Protein Sequence and Structure Codesign
Sep 3, 2026Jiarui Lu, Yuyang Wang, Yizhe Zhang +4Unified Multimodal Models Generative Modeling
Mila, Université de Montréal, Apple · Apple
RosettaSearch: Multi-Objective Inference-Time Search for Protein Sequence Design
Apr 19, 2026Meghana Kshirsagar, Allen Nie, Ching-An Cheng +5Inference-Time Search Large Language Model-Guided Optimization
AI for Good, Microsoft Redmond · Google DeepMind · Google Research +3

Jul 26, 2026J/K move · Enter open · S save

Chamaileon: Cross-Context Binder Design with Contextualized Modeling and Mixed Sampling

Hengyuan Cao, Shizhuo Cheng, Mingxuan Liu, Weicheng Huang+4

Zhejiang University, Hangzhou, China · Fudan University, Shanghai, China · Shanghai Institute for Advanced Study Zhejiang University, Shanghai, China+1

Molecular Generation Protein Design Conditional Generative Modeling Protein Binder Design

Open paper AskPDF arXiv

The rapid evolution of generative models has unlocked new potentials in protein binder design, a pivotal task in structural biology, by facilitating end-to-end generation via joint sequence-structure modeling or hallucination. However, existing approaches are predominantly implemented under a single-target, single-state assumption, limiting their ability to model multi-target or multi-state interactions required for advanced function-oriented protein design. Here, we introduce Chamaileon, which unifies multi-target and multi-state binder design by formulating the problem as cross-context binding landscape modeling. The framework is underpinned by a training paradigm termed In-Context Complex Co-Design (I3CD) for context-aware sequence-structure co-modeling. During inference, we employ Mixture-of-Paths Sampling (MoPS), a scalable strategy that optimizes a single sequence across contexts while alleviating the scarcity of high-quality multi-conformational paired data. Extensive evaluation on our newly constructed benchmark, CROSS, demonstrates that Chamaileon effectively generates sequences adaptable to diverse conformational landscapes and multi-target requirements. The code is available on https://github.com/caohengyuan/Chamaileon.

Abstract

Explore similar work

Chamaileon: Cross-Context Binder Design with Contextualized Modeling and Mixed Sampling

SimpleDesign: A Joint Model for Protein Sequence and Structure Codesign

RosettaSearch: Multi-Objective Inference-Time Search for Protein Sequence Design