通过预测和可解释的机器学习模型,加速针对世卫组织优先病原体的抗微生物发现
Cheng-Ting Tsai1, Chia-Wei Lin1, Gen-Lin Ye1
1Department of Chemistry, National Central University, No. 300, Zhongda Road, Zhongli District, Taoyuan 32001, Taiwan.
ACS omega
|March 4, 2024
概括
创新的机器学习模型可以识别针对耐药病原体的强效抗微生物 (AMP). 这些预测工具,考虑到血液溶解和3D结构,加速新AMP的发现,为传统抗生素提供了重要的替代品.
科学领域:
- 计算化学是一种计算化学.
- 生物技术是生物技术.
- 机器学习在药物发现中的应用.
背景情况:
- 多药耐药 (MDR) 病原体构成了全球健康的重大威胁,推动了对新型抗菌剂的需求.
- 抗微生物 (AMP) 是传统抗生素的有希望的替代品,因为它们的作用机制多样化.
- 传统的AMP发现通常是劳动密集型和昂贵的,需要更有效的识别策略.
研究的目的:
- 开发和验证可预测,可解释的机器学习 (ML) 模型,用于识别强效抗微生物 (AMP).
- 针对世界卫生组织 (WHO) 高优先病原体的AMP,并评估其溶血活性的治疗潜力.
- 加速新型AMP的发现,以应对不断升级的抗生素耐药性.
主要方法:
- 采用了"in silico"方法,专注于从AMP的3D螺旋形状中获得的物理化学属性.
- 开发和应用了尖端的机器学习 (ML) 模型,包括血液溶解预测.
- 采用Shapley添加式解释 (SHAP) 值来解释ML模型的结果并了解作用机制.
主要成果:
- 实现了超过75%的预测准确性,用于识别针对已知和新序列的有力AMP.
- 与本地PEM-2相比,鉴定了几种具有优越抗菌活性的新型AMP,用于对抗WHO优先病原体.
- 证明了ML建模方法在优先考虑和验证有效的AMP候选人的稳健性.
结论:
- 最先进的ML模型可以显著加快新抗菌的设计和发现.
- 开发的预测工具提供了一个强大的策略,通过识别有效的AMP来打击抗生素耐药性.
- 公开可用的预测工具 (https://ai-meta.chem.ncu.edu.tw/amp-meta) 促进了AMP发现的更广泛的研究.
相关概念视频
Antimicrobial Proteins
988
Antimicrobial proteins are important components of the immune system. They aid the body in combating pathogens by either killing them directly or hindering their replication processes. Four main types of antimicrobial substances are interferons, the complement system, iron-binding proteins, and antimicrobial proteins.
Interferons
Interferons (IFNs) are proteins produced by lymphocytes, macrophages, and fibroblasts infected with viruses. While IFNs cannot prevent viruses from entering and...
Interferons
Interferons (IFNs) are proteins produced by lymphocytes, macrophages, and fibroblasts infected with viruses. While IFNs cannot prevent viruses from entering and...
988
Steps in Outbreak Investigation
127
In the ever-evolving field of public health, statistical analysis serves as a cornerstone for understanding and managing disease outbreaks. By leveraging various statistical tools, health professionals can predict potential outbreaks, analyze ongoing situations, and devise effective responses to mitigate impact. For that to happen, there are a few possible stages of the analysis:
127


