生物信息学
计算生物学
毒性
优先次序
肽
数量结构-活动关系
预测值
机器学习
化学
抗菌肽
训练集
一般化
急性毒性
生物
人工智能
计算机科学
化学信息学
生物化学
风险评估
变压器
生物技术
生化工程
药理学
抗菌剂
毒理
作者
Chi‐Ching Lee,Tolulope Joshua Ashaolu
标识
DOI:10.1021/acs.jafc.6c13356
摘要
Peptides from plant sources have antihypertensive, antioxidative, and antimicrobial properties but also entail a risk of allergic and toxic reactions that cannot be detected in conventional wet-lab assays. This review examines deep learning techniques and protein language models for their prediction, from linear motifs screening to transformers (BERT, ESM). The best-performing tools achieve high sensitivity (pLM4Alg 96.2%, ToxinPred3 95.8%) but low specificity and positive predictive value (pLM4Alg 92.8% and 81.7%), meaning that most predictions are false positives. There are four major constraints on the generalization capabilities of these tools: scoring of native peptides while proteolytic processing yields post-translational modifications and neoantigens; training samples correspond to IgE binding, not clinical manifestations, and are biased toward well-characterized organisms; T-cell immunity, microbiome, and organ-specific toxicity are scarcely modeled; and regulatory agencies do not consider in silico predictions sufficient. AI currently serves best as a prioritization filter, not a replacement for biological assays.
科研通智能强力驱动
Strongly Powered by AbleSci AI