PerturbBridge: Direct Phenotypic Supervision for Cellular Perturbation Generation

Yucheng Yang; Xiaoxiao Sun; Huangqingbo Sun; Elaine Sui; Emma Lundberg; Serena Yeung-Levy

cs.CVOct 12, 2026

PerturbBridge: Direct Phenotypic Supervision for Cellular Perturbation Generation

Authors: Yucheng Yang, Xiaoxiao Sun, Huangqingbo Sun, Elaine Sui, Emma Lundberg, Serena Yeung-Levy

Organizations: Stanford University

Abstract

Virtual cell models should capture phenotypes induced by specific perturbations. However, image-based models are typically trained and evaluated to match real-image distributions, which alone does not explicitly prioritize perturbation-specific biological responses. Recent approaches add biological signals through reinforcement-learning post-training or inference-time steering, raising a central question: where should biological response information enter the generative pipeline? We introduce PerturbBridge, a stochastic data-to-data bridge model that directly supervises perturbation-specific phenotypes in predicted perturbed images during training. PerturbBridge jointly optimizes bridge regression and differentiable biological response losses, with biological gradients passing through a frozen image encoder. We further introduce perturbation-response evaluations based on downstream classification and retrieval. Across three microscopy benchmarks for perturbation profiling and subcellular imaging, PerturbBridge surpasses CellFlux in biological fidelity, image quality, and sampling efficiency. On BBBC021, mechanism-of-action classification improves by 5.66 and 7.5 percentage points with morphological and learned visual features, respectively. On JUMP and AllenCell, retrieval Hit@10 and Hit@1 improve by 1.77 and 7.15 percentage points, respectively. On BBBC021 and JUMP, PerturbBridge achieves lower FID with 80% fewer model evaluations than CellFlux. Overall, these results show that directly supervising perturbation responses during training improves phenotypic fidelity while enabling efficient cellular image generation. Code is available at https://github.com/yyc1215lll-creator/PerturbBridge.

Figures & tables

Figure 1: Three mechanisms for incorporating biological response signals. (a) CellFluxRL uses biological rewards for RL post-training. (b) SimuSDB uses phenotypic costs to select candidate transitions during sampling. (c) PerturbBridge jointly optimizes bridge regression and biological response losses on predicted perturbed images during generative training. Conditioning inputs and optional test-time selection in CellFluxRL are omitted for clarity.

Figure 2: Training framework of PerturbBridge. (a) A stochastic data-to-data bridge connects the control image $x_{T}$ at $t=1$ to the perturbed image $x_{0}$ at $t=0$ , with noise strongest at the midpoint and vanishing at both ends. (b) Conditioned on the bridge state $x_{t}$ , perturbation $c$ , and time $t$ , the generative model predicts the perturbed image $\hat{x}_{0}$ . A frozen image encoder extracts features for MoA contrast and condition-moment matching. These objectives encourage separation across MoA classes and align condition-specific feature means and variances with real perturbed data. Biological gradients pass through the encoder to update the generative model jointly with bridge regression.

	CellProfiler features			DINOv2-L/14
Method	Acc. $\uparrow$	Macro-F1 $\uparrow$	W.-F1 $\uparrow$	Acc. $\uparrow$	Macro-F1 $\uparrow$	W.-F1 $\uparrow$
IMPA ( Palma et al., 2025 )	57.9	53.9	60.7	57.8	36.2	60.6
CellFlux ( Zhang et al., 2025 )	72.2	65.8	73.2	67.4	43.4	66.6
CellFlux+bio	73.3	66.6	74.1	69.8	45.4	69.1
Bridge-only	73.1	67.6	74.9	70.6	45.8	70.1
PerturbBridge	77.8	73.0	79.3	74.9	46.9	74.3

Table 1: BBBC021 MoA with CellProfiler features and DINOv2-L/14 features. Bridge-only omits biological response losses. All scores are percentages. Real perturbed data provide a reference. The CellProfiler-based classifier is trained on real perturbed data and tested on 5,120 generated images.

Figure 3: MoA class separation on BBBC021. (a) Real perturbed data; (b) CellFlux; (c) PerturbBridge (ours). Each point represents a cell image, colored by its MoA label. All panels use a shared UMAP projection of the 512-dimensional hidden features from a frozen MoA classifier. Ellipses highlight overlap across MoA classes in CellFlux and clearer separation in PerturbBridge.

	JUMP						AllenCell
Method	Hit @1 $\uparrow$	Hit @5 $\uparrow$	Hit @10 $\uparrow$	MRR $\times 100\uparrow$	mAP $\times 100\uparrow$	Median rank $\downarrow$	Hit @1 $\uparrow$	MRR $\uparrow$	mAP $\uparrow$	Mean rank $\downarrow$
IMPA ( Palma et al., 2025 )	1.50	5.41	9.52	4.65	0.88	84	52.14	0.73	0.73	1.64
CellFlux ( Zhang et al., 2025 )	1.56	5.78	10.38	4.92	0.92	82	62.14	0.78	0.78	1.58
Bridge-only	1.75	5.98	10.42	5.08	0.92	79	62.14	0.79	0.79	1.54
PerturbBridge	2.04	7.27	12.15	5.94	0.97	74	69.29	0.82	0.82	1.48
Real Perturbed Data	2.93	7.94	12.14	6.64	1.10	77	76.19	0.87	0.87	1.29

Table 2: Perturbation retrieval with DINOv2 features. JUMP uses single-cell Mahalanobis retrieval across plates. AllenCell uses condition profiles to retrieve compounds within the same target from profiles of real perturbed training data. Bridge-only omits biological response losses and corresponds to the ECSI-only baseline on AllenCell. Hit rates are percentages; bold and underlined values mark the best and second-best generated-image results within each panel.

Figure 4: Comparison of perturbation-induced changes on AllenCell. Columns show seven compound–structure combinations involving the endoplasmic reticulum (ER), myosin bundles, actin filaments, tight junctions, and Golgi apparatus. Control images appear at the top, PerturbBridge samples in the middle (Ours), and real perturbed images at the bottom (Real). Cell nuclei are shown in purple and plasma membranes in blue . SNB denotes (S)-nitro-blebbistatin; annotations highlight subcellular changes. Examples are shown in App. F .

	Standard split			Unseen compounds
Method	Acc. $\uparrow$	Macro-F1 $\uparrow$	W.-F1 $\uparrow$	Acc. $\uparrow$	Macro-F1 $\uparrow$	W.-F1 $\uparrow$
PhenDiff ( Bourou et al., 2024 )	52.6	33.6	52.1	9.6	9.3	7.4
IMPA ( Palma et al., 2025 )	63.7	40.2	64.8	16.0	10.0	13.1
CellFlux ( Zhang et al., 2025 )	64.9	42.1	63.8	43.2	36.6	42.8
CellFluxRL ( Wu et al., 2026 )	66.2	42.9	67.9	–	–	–
CellFlux+bio	66.3	42.6	65.1	44.7	39.5	45.2

Table 3: BBBC021 MoA with Inception-v3 features on the standard and unseen-compound splits. Classification scores are percentages; bold and underlined values mark the best and second-best generated-image results within each split. Dashes indicate unavailable results. Bridge-only omits biological response losses. CellFlux+bio applies matched biological response losses to the predicted perturbed image reconstructed from velocity outputs. Gray rows denote ablation variants.

	BBBC021				JUMP				AllenCell
Method	FID $\downarrow$	KID $\downarrow$	FDD $\downarrow$	KDD $\downarrow$	FID $\downarrow$	KID $\downarrow$	FDD $\downarrow$	KDD $\downarrow$	FID $\downarrow$	KID $\downarrow$	FDD $\downarrow$	KDD $\downarrow$
PhenDiff ( Bourou et al., 2024 )	49.5	3.10	250.1	1.55	49.3	5.10	247.3	2.76	48.8	2.75	192.6	1.56
IMPA ( Palma et al., 2025 )	33.7	2.60	293.9	2.57	14.6	1.10	371.1	4.43	37.6	2.25	424.1	4.98
CellFlux ( Zhang et al., 2025 )	18.8	1.61	207.2	1.73	9.0	0.63	123.1	1.35	28.4	1.46	181.2	1.71
CellFluxRL ( Wu et al., 2026 )	24.0	1.40	–	–	–	–	–	–	–	–	–	–
SimuSDB ( Zhang et al., 2026a )	16.5	1.32	–	–	8.7	0.58	–	–	–	–	–	–

Table 4: Image-distribution results in Inception-v3 (FID, KID) and DINOv2 (FDD, KDD) feature spaces. Lower is better. KID is multiplied by 100; KDD is unscaled. Bold and underlined values mark the best and second-best results within each dataset and metric. Dashes denote unreported values. CellFluxRL and SimuSDB FID/KID values are from Wu et al. (2026) and Zhang et al. (2026a) , respectively. Gray rows denote ablation variants.

Figure 5: Sampling-budget comparison on BBBC021 and JUMP. Lower FID is better. Gray bars: CellFlux; blue bars: PerturbBridge. Bar labels report FID; horizontal axes indicate model evaluations (NFE).

Appendix figures & tables8 assets

Supplementary material from the paper’s appendix.

Appendix

Figure 6: Image fidelity and biological correctness. Orange fibers depict the target phenotype; blue-gray texture depicts non-target intracellular appearance. Paired enlargements isolate both features. Relative to (a), (b) matches non-target appearance but disrupts the phenotype, whereas (c) preserves the phenotype despite a texture shift. Population-feature ellipses schematize the resulting disagreement between FID and biological correctness; the FID ordering follows Proposition 1 . Cells and ellipses are conceptual illustrations.

Component	Setting
Backbone	56,201,219-parameter ADM/U-Net; RGB $96\times 96$
Architecture	Base width 128; four residual blocks; multipliers $(2,2,2)$ ; attention resolution 48; one head; dropout 0.3
Blocks	Scale–shift normalization; new attention order; convolutional resampling and residual up/downsampling disabled
Perturbation	1,024-dimensional molecular embedding; local-batch condition dropout 0.2
Batching	Eight devices, 16 cells each; eight cells per compound; global batch 128
Optimizer	RAdam; learning rate $10^{-4}$ ; weight decay 0; default betas; no learning-rate annealing

Appendix

Table 5: BBBC021 training configuration.

Metric	BBBC021	JUMP	AllenCell
FID	0.493	0.230	0.530
KID $\times 100$	0.072	0.039	0.126
FDD	6.855	3.410	7.940
KDD	0.089	0.065	0.095
MoA Accuracy (p.p.)	0.350	—	—
MoA Macro-F1 (p.p.)	0.887	—	—

Appendix

Table 6: Standard deviations across three random seeds for PerturbBridge on BBBC021, JUMP, and AllenCell; reported results are three-seed means. KID is multiplied by 100. MRR and mAP are multiplied by 100 on JUMP and remain unscaled on AllenCell, matching Table 2 . MoA accuracy, F1 scores, and Hit@1 standard deviations are expressed in percentage points (p.p.). Standard deviations are rounded to three decimal places; dashes indicate unavailable statistics.

Figure 7: Drug-induced changes across seven cellular structures on AllenCell. PerturbBridge samples (left) and real references (right) are arranged by compound and structure. Vehicle denotes the control; SNB denotes (S)-nitro-blebbistatin. Yellow frames highlight selected comparisons; blank entries denote combinations absent from the dataset.

Figure 8: BBBC021 examples for cytochalasin B and AZ138. In each compound block, the left column contains real perturbed reference images and the remaining columns contain PerturbBridge samples.

Figure 9: Additional BBBC021 examples for AZ258 and PP-2. Real perturbed references appear in the left column and PerturbBridge samples in the remaining columns.

Figure 10: AllenCell examples for brefeldin A (Golgi) and paclitaxel (microtubules). The left column shows real vehicle and treated references; the remaining columns show PerturbBridge samples.

Figure 11: AllenCell examples for (S)-nitro-blebbistatin (SNB) under Golgi and tight-junction imaging. Real vehicle and treated references appear in the left column; PerturbBridge samples occupy the remaining columns.

Single-Cell Perturbation Modeling

Single-cell perturbation modeling predicts how genetic or chemical interventions change cellular states across cells and contexts.

See all 58 papers

Jun 9, 2026

FreeBridge: Variational Schrödinger Bridges for Cellular Transition Dynamics

High-content imaging assays quantify cellular responses to chemical and genetic perturbations, yet continuous trajectories of individual cells are unobservable because cells are chemically fixed at acquisition. Perturbation modeling therefore reduces to inferring stochastic transport between control and treated populations observed only as separate marginals. While recent generative models achieve strong end-point alignment, boundary consistency does not determine intermediate evolution: multiple stochastic processes may connect identical marginals while traversing regions unsupported by observed single-cell morphologies. We introduce FreeBridge, a Schrödinger Bridge formulation for single-cell transition modeling under endpoint-only supervision. FreeBridge defines atomic states as instance-segmented single-cell representations, establishing a fixed cellular manifold, and learns stochastic transport constrained within this geometry via empirical latent support regularization. Across BBBC021, RxRx1, and JUMP, FreeBridge maintains competitive or improved endpoint fidelity and mechanism-of-action retention under a unified evaluation protocol; on BBBC021, it further reduces intermediate support violations. These findings highlight the importance of geometric grounding for biologically interpretable perturbation dynamics. Project page: https://y-research-sbu.github.io/FreeBridge/.

Xurui Wang, Qin Ren, Jun Ma +2

Stony Brook University, Stony Brook, NY, USA · University of Toronto, Toronto, ON, Canada · University Health Network, Toronto, ON, Canada

Schrödinger Bridges Single-Cell Perturbation Modeling Microscopy Image Analysis

Ask PDF

Feb 23, 2026

PerturbDiff: Functional Diffusion for Single-Cell Perturbation Modeling

Building Virtual Cells that can accurately simulate cellular responses to perturbations is a long-standing goal in systems biology. A fundamental challenge is that high-throughput single-cell sequencing is destructive: the same cell cannot be observed both before and after a perturbation. Thus, perturbation prediction requires mapping unpaired control and perturbed populations. Existing models address this by learning maps between distributions, but typically assume a single fixed response distribution when conditioned on observed cellular context (e.g., cell type) and the perturbation type. In reality, responses vary systematically due to unobservable latent factors such as microenvironmental fluctuations and complex batch effects, forming a manifold of possible distributions for the same observed conditions. To account for this variability, we introduce PerturbDiff, which shifts modeling from individual cells to entire distributions. By embedding distributions as points in a Hilbert space, we define a diffusion-based generative process operating directly over probability distributions. This allows PerturbDiff to capture population-level response shifts across hidden factors. Benchmarks on established datasets show that PerturbDiff achieves state-of-the-art performance in single-cell response prediction and generalizes substantially better to unseen perturbations. See our project page (https://katarinayuan.github.io/PerturbDiff-ProjectPage/), where code and data will be made publicly available (https://github.com/DeepGraphLearning/PerturbDiff).

Xinyu Yuan, Xixian Liu, Ya Shi Zhang +3

Mila - Québec AI Institute · University of Montréal · University of Ottawa +3

Single-Cell Perturbation Modeling Single-Cell Transcriptomics Diffusion Models Conditional Diffusion Models

Ask PDF

Sep 17, 2026

CellRFT: Reinforcement Fine-Tuning for Single-Cell Perturbation Modeling

Predicting cellular responses to perturbations supports the study of gene function, disease mechanisms, and therapeutic strategies. Despite advances in single-cell perturbation modeling, existing models typically optimize surrogate losses that do not directly reflect the biological criteria used for evaluation, so better data fitting need not yield better biological predictions. To address this mismatch, we introduce CellRFT, a reinforcement fine-tuning framework that uses biological evaluation as direct training feedback. CellRFT uses policy-gradient optimization to learn from non-differentiable evaluations of generated cell populations and integrates multiple biological rewards through hierarchical reward aggregation. Comprehensive experiments demonstrate CellRFT's applicability across different pretrained models and effectiveness in improving perturbation prediction, reveal that optimizing one biological criterion can help or hinder others, and show that complementary rewards can improve criteria beyond those directly optimized, offering a way to probe how biological metrics shape model behavior, with the potential to inform evaluation design. Code will be made available.

Jie Yan, Li Liu, Hanze Guo +7

Chinese Academy of Sciences · Peking University · Renmin University of China +2

Multi-Objective Reinforcement Learning Single-Cell Perturbation Modeling RL Fine-Tuning

Ask PDF

Abstract

Single-Cell Perturbation Modeling

Explore similar work

FreeBridge: Variational Schrödinger Bridges for Cellular Transition Dynamics

PerturbDiff: Functional Diffusion for Single-Cell Perturbation Modeling

CellRFT: Reinforcement Fine-Tuning for Single-Cell Perturbation Modeling