From passive tools to autonomous partners: the rise of AI agents in bioinformatics
AI Biology & Medicine · 2026-11 · 100002
DOI: 10.1016/j.aibm.2026.100002 ↗Kurz gefasst
Fachartikel in der Zeitschrift AI Biology & Medicine, 2026-11. Verfasst von Jing-Jing Peng und 5 weiteren Autor:innen. Die Arbeit nennt 130 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: National Natural Science Foundation of China.
Abstract (Originalsprache)
Large language model (LLM)-driven agents have emerged as a promising paradigm in bioinformatics, enabling autonomous and adaptive analysis of complex biological problems. This review comprehensively summarizes the fundamental frameworks of artificial intelligence (AI) agents, breaking down their core architecture into essential components: an LLM brain, planning, memory, and tools. Furthermore, we categorize current biomedical AI agents into general agents—designed for broad interdisciplinary tasks—and specialized agents tailored for distinct biological data types. We further organize specialized agents along two dimensions—data modalities and representative applications—with a particular focus on genomics, single-cell transcriptomics, proteomics, spatial omics, and multi-omics analysis, while highlighting drug discovery as a representative application scenario. Despite their potential to automate complex workflows, hurdles such as model hallucinations, “black-box” reasoning, and biosafety risks persist. We highlight the transition toward embodied intelligence and self-driving laboratories, envisioning natively multimodal agents as critical drivers in accelerating the pace of biological discovery.
Förderung
- National Natural Science Foundation of China
Zitieren
APA
Peng, J., Zeng, F., Yang, J., Lin, Y., Lei, Q., & Guo, A. (2026). From passive tools to autonomous partners: the rise of AI agents in bioinformatics. AI Biology & Medicine, 100002. https://doi.org/10.1016/j.aibm.2026.100002
Vancouver
Peng J, Zeng F, Yang J, Lin Y, Lei Q, Guo A. From passive tools to autonomous partners: the rise of AI agents in bioinformatics. AI Biology & Medicine. 2026:100002. doi:10.1016/j.aibm.2026.100002
BibTeX
@article{Peng2026,
author = {Peng, Jing-Jing and Zeng, Fei-Ran and Yang, Jing-Min and Lin, Yu and Lei, Qian and Guo, An-Yuan},
title = {From passive tools to autonomous partners: the rise of AI agents in bioinformatics},
journal = {AI Biology & Medicine},
year = {2026},
pages = {100002},
doi = {10.1016/j.aibm.2026.100002}
}
EUVIMED