Exploring the Potential of Large Language Models in Molecular Tasks: An Insightful Evaluation with GPT-4
Zhang, J.; Fang, Y.; Zhang, N.; Shao, X.; Chen, H.; Fan, X.
Show abstract
In the rapidly changing realm of artificial intelligence, large language models (LLMs) such as GPT-4 are increasingly being explored for their potential to aid and enhance the field of molecular research. This study explores the performance of GPT-4 and GPT-3.5 in molecular research, particularly in generating and optimizing molecular structures. The results highlight GPT-4s strengths in certain areas of molecular optimization, while also revealing challenges in accurately generating complex molecules. The findings underscore the necessity for integrating these models with domain-specific tools to enhance their application in scientific research, particularly in molecular studies. The study offers insights into the potential of LLMs for advancing molecular research, paving the way for future developments in this rapidly evolving field.
Matching journals
The top 4 journals account for 50% of the predicted probability mass.
Similar papers in this journal
- Retro Drug Design: From Target Properties to Molecular Structures 96%
- Streamlining Computational Fragment-Based Drug Discovery through Evolutionary Optimization Informed by Ligand-Based Virtual Prescreening 96%
- Advancements in Ligand-Based Virtual Screening through the Synergistic Integration of Graph Neural Networks and Expert-Crafted Descriptors 95%
Similar papers in this journal
- Chemical Genomics Language Model toward Reliable and Explainable Compound-Protein Interaction Exploration 96%
- PL-PatchSurfer3: Improved Structure-Based Virtual Screening for Structure Variation Using 3D Zernike Descriptors 96%
- Deep learning integration of molecular and interactome data for protein-compound interaction prediction 96%
Similar papers in this journal
- Binding affinity prediction for protein-ligand complex using deep attention mechanism based on intermolecular interactions 96%
- A Comprehensive Survey of Scoring Functions for Protein Docking Models 94%
- vCOMBAT: a Novel Tool to Create and Visualize a COmputational Model of Bacterial Antibiotic Target-binding 94%
"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.