Enhancing Safe and Controllable Protein Generation via Knowledge Preference Optimization

ACL ARR 2024 December Submission877 Authors

Abstract: Protein language models have emerged as powerful tools for sequence generation, offering substantial advantages in functional optimization and $de novo$ design. However, these models also present significant risks of generating harmful protein sequences, such as those that enhance viral transmissibility or evade immune responses. These concerns underscore critical biosafety and ethical challenges. To address these issues, we propose a Knowledge-guided Preference Optimization (KPO) framework that integrates prior knowledge via a Protein Safety Knowledge Graph. This framework utilizes an efficient graph pruning strategy to identify preferred sequences and employs reinforcement learning to minimize the risk of generating harmful proteins. Experimental results demonstrate that KPO effectively reduces the likelihood of producing hazardous sequences while maintaining high functionality, offering a robust safety assurance framework for applying generative models in biotechnology.

Paper Type: Long

Research Area: NLP Applications

Research Area Keywords: healthcare applications; knowledge graphs;

Contribution Types: Publicly available software and/or pre-trained models, Data resources

Languages Studied: English; Protein Languages

Submission Number: 877