Instruction Multi-Constraint Molecular Generation Using a Teacher-Student Large Language Model

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Zhou, Peng, Wang, Jianmin, Li, Chunyan, Wang, Zixu, Liu, Yiping, Sun, Siqi, Lin, Jianxin, Wei, Leyi, Cai, Xibao, Lai, Houtim, Liu, Wei, Wang, Longyue, Liu, Yuansheng, Zeng, Xiangxiang
Format: Preprint
Veröffentlicht: 2024
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866910643066503168
author Zhou, Peng
Wang, Jianmin
Li, Chunyan
Wang, Zixu
Liu, Yiping
Sun, Siqi
Lin, Jianxin
Wei, Leyi
Cai, Xibao
Lai, Houtim
Liu, Wei
Wang, Longyue
Liu, Yuansheng
Zeng, Xiangxiang
author_facet Zhou, Peng
Wang, Jianmin
Li, Chunyan
Wang, Zixu
Liu, Yiping
Sun, Siqi
Lin, Jianxin
Wei, Leyi
Cai, Xibao
Lai, Houtim
Liu, Wei
Wang, Longyue
Liu, Yuansheng
Zeng, Xiangxiang
contents While various models and computational tools have been proposed for structure and property analysis of molecules, generating molecules that conform to all desired structures and properties remains a challenge. Here, we introduce a multi-constraint molecular generation large language model, TSMMG, which, akin to a student, incorporates knowledge from various small models and tools, namely, the 'teachers'. To train TSMMG, we construct a large set of text-molecule pairs by extracting molecular knowledge from these 'teachers', enabling it to generate novel molecules that conform to the descriptions through various text prompts. We experimentally show that TSMMG remarkably performs in generating molecules meeting complex, natural language-described property requirements across two-, three-, and four-constraint tasks, with an average molecular validity of over 99% and success ratio of 82.58%, 68.03%, and 67.48%, respectively. The model also exhibits adaptability through zero-shot testing, creating molecules that satisfy combinations of properties that have not been encountered. It can comprehend text inputs with various language styles, extending beyond the confines of outlined prompts, as confirmed through empirical validation. Additionally, the knowledge distillation feature of TSMMG contributes to the continuous enhancement of small models, while the innovative approach to dataset construction effectively addresses the issues of data scarcity and quality, which positions TSMMG as a promising tool in the domains of drug discovery and materials science.
format Preprint
id arxiv_https___arxiv_org_abs_2403_13244
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Instruction Multi-Constraint Molecular Generation Using a Teacher-Student Large Language Model
Zhou, Peng
Wang, Jianmin
Li, Chunyan
Wang, Zixu
Liu, Yiping
Sun, Siqi
Lin, Jianxin
Wei, Leyi
Cai, Xibao
Lai, Houtim
Liu, Wei
Wang, Longyue
Liu, Yuansheng
Zeng, Xiangxiang
Computation and Language
Artificial Intelligence
While various models and computational tools have been proposed for structure and property analysis of molecules, generating molecules that conform to all desired structures and properties remains a challenge. Here, we introduce a multi-constraint molecular generation large language model, TSMMG, which, akin to a student, incorporates knowledge from various small models and tools, namely, the 'teachers'. To train TSMMG, we construct a large set of text-molecule pairs by extracting molecular knowledge from these 'teachers', enabling it to generate novel molecules that conform to the descriptions through various text prompts. We experimentally show that TSMMG remarkably performs in generating molecules meeting complex, natural language-described property requirements across two-, three-, and four-constraint tasks, with an average molecular validity of over 99% and success ratio of 82.58%, 68.03%, and 67.48%, respectively. The model also exhibits adaptability through zero-shot testing, creating molecules that satisfy combinations of properties that have not been encountered. It can comprehend text inputs with various language styles, extending beyond the confines of outlined prompts, as confirmed through empirical validation. Additionally, the knowledge distillation feature of TSMMG contributes to the continuous enhancement of small models, while the innovative approach to dataset construction effectively addresses the issues of data scarcity and quality, which positions TSMMG as a promising tool in the domains of drug discovery and materials science.
title Instruction Multi-Constraint Molecular Generation Using a Teacher-Student Large Language Model
topic Computation and Language
Artificial Intelligence
url https://arxiv.org/abs/2403.13244