A GPT-2 model fine-tuned with disease MeSH + protein sequence inputs and GRPO rewards produces more unique, valid, drug-like, high-PLAPT-affinity ligands than DrugGPT or DrugGen on five diabetic-nephropathy targets.
Phenotypic drug discovery: recent successes, lessons learned and new direc tions
2 Pith papers cite this work, alongside 322 external citations. Polarity classification is still indexing.
2
Pith papers citing it
322
external citations · OpenAlex
citation-role summary
background 1
citation-polarity summary
years
2026 2roles
background 1polarities
background 1representative citing papers
DINO-based ViT models pretrained on HPA FOV achieve macro F1 of 0.822 zero-shot and 0.860 after fine-tuning for protein localization on OpenCell, demonstrating effective transfer from SSL pretraining.
citing papers explorer
-
DrugGen 2: A disease-aware language model for enhancing drug discovery
A GPT-2 model fine-tuned with disease MeSH + protein sequence inputs and GRPO rewards produces more unique, valid, drug-like, high-PLAPT-affinity ligands than DrugGPT or DrugGen on five diabetic-nephropathy targets.
-
Using Deep Learning Models Pretrained by Self-Supervised Learning for Protein Localization
DINO-based ViT models pretrained on HPA FOV achieve macro F1 of 0.822 zero-shot and 0.860 after fine-tuning for protein localization on OpenCell, demonstrating effective transfer from SSL pretraining.