Unseen Perturbation Prediction#
đź’ˇ Tip: Click the + in the first column to expand detailed method descriptions.
| Method | Year | Task | Model | Published | Code | |
|---|---|---|---|---|---|---|
| AttentionPert | 2024 | Combinatorial Effect PredictionUnseen Perturbation Prediction |
Transformer ModelGNN |
âś“ | GitHub | |
| Biolord | 2024 | Multi-component DisentanglementUnseen Perturbation PredictionCombinatorial Effect PredictionContext Transfer |
ProbabilisticZINB likelihoodProtein-Count (totalVI) Likelihood |
âś“ | GitHub | |
| C2S-Scale | 2025 | Unseen Perturbation PredictionCombinatorial Effect Prediction |
Family of LLMs with to 27B parameters |
âś— | GitHub | |
| cellFlow | 2025 | Trace Cell PopulationsContext TransferUnseen Perturbation PredictionCombinatorial Effect Prediction |
Conditional Flow MatchingOptimal Transport |
âś— | GitHub | |
| ChemCPA | 2022 | Unseen Perturbation PredictionContext TransferMulti-component Disentanglement |
AE framework inspired by CPAChemical representation embeddings |
âś“ | GitHub | |
| CODEX | 2024 | Combinatorial Effect PredictionUnseen Perturbation PredictionContext Transfer |
DNN |
âś“ | GitHub | |
| CondOT | 2022 | Trace Cell PopulationsUnseen Perturbation PredictionCombinatorial Effect PredictionContext Transfer |
Conditioned Dual (min-max) Formulation OT |
âś“ | GitHub | |
| cycleCDR | 2024 | Unseen Perturbation Prediction |
Autoencoder |
âś“ | GitHub | |
| GEARS | 2023 | Combinatorial Effect PredictionUnseen Perturbation Prediction |
GNNs for co-expression and GO relationshipsLabel embeddingsPK Representations |
âś“ | GitHub | |
| GeneCompass | 2024 | Unseen Perturbation PredictionCombinatorial Effect PredictionGRN InferenceNonlinear Gene Programmes |
Foundational Gene expression embeddings (from >50M human cells)Self-supervised masked regression with down-samplingSparse transformer encoderPerformer-style attention decoderPK-informed |
âś“ | GitHub | |
| IterPert | 2024 | Unseen Perturbation PredictionCombinatorial Effect Prediction |
GEARSActive Learning |
âś“ | GitHub | |
| LLM+GP | 2024 | Unseen Perturbation Prediction |
Gaussian Process ModelLanguage embeddings |
âś“ | âś— | |
| LPM | 2025 | Unseen Perturbation PredictionCombinatorial Effect Prediction |
DNN Decoder |
âś— | âś— | |
| PDGrapher | 2025 | Combinatorial Effect PredictionUnseen Perturbation PredictionContext Transfer |
GNN |
âś“ | GitHub | |
| PRNet | 2024 | Unseen Perturbation Prediction |
DNN |
âś“ | GitHub | |
| Prophet | 2024 | Unseen Perturbation PredictionContext Transfer |
TransformerMulti-modalKnowledge Informed |
âś— | GitHub | |
| SCCVAE | 2025 | Unseen Perturbation PredictionCausal Structure |
VAEVariational InferenceLatent Causal Structure |
âś— | GitHub | |
| scFoundation | 2024 | Nonlinear Gene ProgrammesUnseen Perturbation PredictionCombinatorial Effect Prediction |
Foundational Gene expression embeddings (from >50M human cells)Self-supervised masked regression with down-samplingSparse transformer encoderPerformer-style attention decoder |
âś“ | GitHub | |
| scGenePT | 2025 | Unseen Perturbation PredictionCombinatorial Effect PredictionGRN InferenceNonlinear Gene Programmes |
scGPTChatGPT prompts |
âś— | âś— | |
| scGPT | 2024 | Unseen Perturbation PredictionCombinatorial Effect PredictionGRN InferenceNonlinear Gene Programmes |
Foundational Gene expression embeddings (from >33M human cells)Self-supervised masked expression predictionCustomised non-sequential (flash) attention |
âś“ | GitHub | |
| Squidiff | 2025 | Combinatorial Effect PredictionContext TransferUnseen Perturbation Prediction |
Diffusion Model |
âś“ | GitHub |