MOLGEN

← Back to all benchmarks

Date: 2024-12-17

Name: MOLGEN

Domain: Chemistry

Focus: Molecular generation and optimization

Task Types: Distribution learning, Goal-oriented generation

Metrics: Validity%, Novelty%, QED, Docking score, penalized logP

Models: MolGen

AI/ML Motif: Generative

Resources

Benchmark: Visit

Datasets: MolGen: A Pre-trained Molecular Language Model

Results: Domain-Agnostic Molecular Generation with Chemical Feedback

Keywords

SELFIES GAN property optimization

Citation

Yin Fang, Ningyu Zhang, Zhuo Chen, Lingbing Guo, Xiaohui Fan, and Huajun Chen. Domain-agnostic molecular generation with chemical feedback. 2024. URL: https://arxiv.org/abs/2301.11259, arXiv:2301.11259.

@misc{fang2024domainagnosticmoleculargenerationchemical,
  archiveprefix = {arXiv},
  author        = {Yin Fang and Ningyu Zhang and Zhuo Chen and Lingbing Guo and Xiaohui Fan and Huajun Chen},
  eprint        = {2301.11259},
  primaryclass  = {cs.LG},
  title         = {Domain-Agnostic Molecular Generation with Chemical Feedback},
  url           = {https://arxiv.org/abs/2301.11259},
  year          = {2024}
}

Ratings

CategoryRating

Software

5.00

Code is available on the github repo, along with instructions to run the model and reproduce results.

Specification

4.00

Task, datset format, and input/output formats are well specified. No system constraints are mentioned.

Dataset

5.00

Dataset and train/test splits are available through the github repo, as well as mentions of source datasets in the paper.

Metrics

5.00

Metrics are well defined and appropriate for the task

Reference Solution

5.00

A pretrained model is provided, as well as training code and instructions

Documentation

5.00

All necessary information is provided in the paper and github repo

Average rating: 4.83/5

Radar plot

$MOLGEN radar$

Edit: edit this entry