liveBook · Manning

Geneformer

As explained in Build AI Drug Discovery Pipelines

Geneformer is a transcriptome-level model that learns gene co-expression patterns from single-cell RNA sequencing data.

Geneformer analyzes single-cell RNA sequencing data at the transcriptome level: it learns patterns in how genes are expressed together across cell types and how expression changes with drug treatment or disease. This broader view helps place a potential drug target within its regulatory network. section 13.2.1

Prioritizing a drug target

For a candidate target, Geneformer can mask the target gene in single-cell transcriptomic profiles and predict changes in cellular state. The result is a computational hypothesis, not a validated knockout phenotype. It can help prioritize targets and flag potential liabilities for experimental follow-up. section 13.2.1

How Build AI Drug Discovery Pipelines covers it

How deep it goes
Explains the model’s transcriptome-level role and a masking-based hypothesis workflow, without implementation details.
Examples
Uses drug treatment, disease states, and a masked candidate gene to illustrate expression changes and predicted knockout effects.
What you can do after
Evaluate and understand computational knockout hypotheses as priorities for experimental follow-up rather than validated phenotypes.
How it connects
Connects single-cell RNA sequencing and gene co-expression to regulatory networks and target validation.

This explainer was generated by Manning's knowledge graph from the text of Build AI Drug Discovery Pipelines. The section links point to the parts of the book it draws on.