bio.rodeo
ModelsOrganizationsProvidersLeaderboardAboutSign in
bio.rodeo

The authoritative source for evaluating biological foundation models. No hype, just honest analysis.

Categories
  • DNA & Gene
  • RNA
  • Protein
  • Small molecule
  • Single-cell
  • Spatial omics
  • Pathology
  • Imaging
  • Metabolomics
  • Biosignals
  • Language model
bio.rodeoModelsOrganizationsProvidersLeaderboardAboutFAQSubmit a modelTermsPrivacyContact
© 2026 Pulsatance. All rights reserved. ~
Built by Pulsatance
models / single-cell / instructcell
Single-cellLanguage model
Zhejiang UniversityReleased January 2025

InstructCell

Single-cell analysis model driven by plain-language instructions, covering cell type annotation, pseudo-cell generation, and drug response prediction.

The short version

  • —Answers annotation, generation and drug-response questions typed as ordinary sentences
  • —A Q-Former compresses each expression profile into eight tokens the language model reads
  • —Writes cells back out as ZINB gene counts through a conditional VAE decoder
  • —Trained on human and mouse profiles jointly, so one checkpoint spans both species
  • —Chat variant narrates its reasoning; instruct variant returns just the answer
79Openness

Where to run it

No providers recorded yet. Browse all providers

Single-cell RNA sequencing produces data that most analysis pipelines expose only through code: annotating cell types, simulating a cell population, or asking whether a cell will respond to a drug each means assembling a bespoke script around a task-specific model. InstructCell removes that layer by making natural language the interface. A user types an ordinary sentence alongside an expression profile, and the model decides which task is being asked for and returns either a textual answer or a synthesised cell.

InstructCell was developed at Zhejiang University by the group behind ChatCell, and released as a preprint in January 2025. Where ChatCell converted cells into sentences of gene names for a text-only model, InstructCell keeps gene expression in its native numeric form and gives the language model a dedicated encoder and decoder for it. That makes the transcriptome a first-class modality rather than something serialised into text, and it lets the same checkpoint both read cells and write them.

The model sits alongside single-cell foundation models such as scGPT, Geneformer, and scBERT, but occupies a different position: those models supply general-purpose cell embeddings that a practitioner adapts per task, whereas InstructCell packages three concrete analyses behind one conversational entry point.

#Key Features

  • Instruction-driven task dispatch: Cell type annotation, conditional pseudo-cell generation, and drug sensitivity prediction are all invoked by phrasing a request in plain English, with no task flag, adapter swap, or per-task head to select.
  • Gene expression as a native modality: A Q-Former encoder converts a raw expression vector into a small set of tokens the language model consumes directly, avoiding the information loss of rank-ordering genes into text.
  • Generative cell output: A conditional variational autoencoder decodes the language model's hidden state back into gene counts under a zero-inflated negative binomial likelihood, matching the sparsity and overdispersion of real scRNA-seq measurements.
  • Robustness to how a question is asked: Instruction templates were synthesised across a wide range of communication styles and expertise levels, so the model tolerates the phrasing variability of real users rather than a fixed command vocabulary.
  • Two released variants: A chat version that produces context-rich prose alongside its output, and an instruct version tuned for terse, task-focused answers.

#Technical Details

InstructCell uses a T5-base encoder-decoder as its language backbone. Expression profiles enter through a Q-Former module with four transformer layers, eight learnable query tokens, and six key-value pairs, feeding a three-layer MLP with skip connections; ablations show eight query tokens to be optimal, with additional queries introducing redundancy. Cell generation runs through a conditional VAE whose encoder produces posteriors over a latent and a library-size term, and whose decoder emits zero-inflated negative binomial parameters over genes; a fully connected layer projects the 768-dimensional hidden state to the 256-dimensional conditioning vector that drives it.

Training data comprises 299,155 scRNA-seq profiles drawn from ten public datasets covering human and mouse tissues including pancreas, blood, lung, and bone marrow. Mouse genes are mapped to human orthologs, and the top 3,600 highly variable genes per dataset are merged into a shared vocabulary of 18,961 genes. Instruction-response templates were synthesised with GPT-4o across 79 personality traits and per-task motivations, deduplicated at a ROUGE-L threshold of 0.75, yielding 2,787 templates for annotation, 2,395 for drug sensitivity, and 2,080 for pseudo-cell generation. On annotation the model matches or exceeds scBERT, scGPT, and Geneformer on weighted F1, macro F1, and accuracy; on generation it achieves lower maximum mean discrepancy than scDiffusion and scGAN; on drug sensitivity it leads on all three metrics for the GSE117872 and GSE110894 cohorts and is comparable on GSE149383.

#Applications

The intended user is a bench biologist or clinician-scientist who has scRNA-seq data and a question, but not a computational pipeline. Annotating an unlabelled dataset, generating synthetic cells of a specified type to augment a rare population, and screening predicted drug response across a tumour cell population all become single-sentence requests against a released checkpoint. Because the gene vocabulary spans human and mouse orthologs, the same model serves cross-species comparisons without a separate checkpoint per organism.

#Impact

InstructCell is an argument that the bottleneck in single-cell analysis is increasingly the interface rather than the representation: general-purpose cell embeddings are plentiful, but turning one into an answer still demands expertise most wet-lab groups do not have. By pairing a modest 220M-parameter backbone with purpose-built encoders and decoders for expression data, it reaches parity with much larger single-cell foundation models on the tasks it covers, which suggests that instruction tuning on a well-constructed multi-modal dataset can substitute for raw scale. Its coverage is deliberately narrow — three tasks, ten training datasets — and the work remains a preprint awaiting peer review. Code, both checkpoints, and the preprocessing pipeline are released under the MIT license.

At a glance

Released
January 2025
Category
Single-cell
Organization
Zhejiang University

Links

GitHub RepositoryResearch PaperHuggingFace ModelHuggingFace Model

Tags

cell_type_annotationdrug_sensitivity_predictioninstruction_tuningmultimodaltransformervariational_autoencoder

Something wrong?

Much of this page is generated or calculated automatically. Flag anything that looks off and we will re-run it.