Published February 23, 2026 | Version v1
Model Open

Phenotype-Guided In Silico Molecular Generation Using Large Language Models

Authors/Creators

Description

GEMGen is a generative model for phenotype-based drug discovery.
It consists of two core components:

  • Generator: generates small-molecule candidates conditioned on cell type information and lists of up- and down-regulated genes from differential expression analysis.
  • Scorer: evaluates how well a generated small molecule matches the given phenotypic input.

The framework is designed for flexible inference and can be easily adapted to different biological contexts and datasets. Users need to download this model folder and use it in conjunction with the scripts from https://github.com/DLS5-Omics/GEMGen.

Files

models.zip

Files (7.7 GB)

Name Size
md5:acdb315c6097c095cd5f4f8a5bd87543
7.7 GB Preview Download

Additional details

Identifiers

DOI
10.64898/2026.01.03.697483

Dates

Submitted
2026-02-23

Software

Repository URL
https://github.com/DLS5-Omics/GEMGen
Programming language
Python
Development Status
Active