PromptBio: An Agentic Platform for End-to-End Computational Biomedical Research
- 1PromptBio Inc
Abstract
Modern biomedical research increasingly depends on complex computational analyses, yet translating a scientific question into a reliable workflow still requires substantial technical expertise and manual coordination. PromptBio is a multi-agent AI platform available through a web portal at https://promptbio.ai that addresses this challenge. Through natural-language interaction, its agent harness, PromptGenie, translates a research objective into an inspectable plan, executes the required research and analysis, and adapts subsequent steps as evidence and results emerge. PromptBio integrates reasoning with managed execution while preserving human oversight and a traceable record of the research process. It can apply validated methods, construct custom analyses, and incorporate external workflows, enabling researchers to move beyond fixed pipelines while supporting reproducibility. We evaluate the platform through benchmarks of end-to-end bioinformatics analysis and biomedical deep research, validation of representative omics and machine-learning skills, and a hypothesis-driven regulatory-genomics case study. PromptGenie achieved higher analytical accuracy and stronger evidence retrieval and synthesis than the comparison agents, while the evaluated domain skills produced results consistent with established methods. The case study further demonstrates how PromptBio can integrate human-specific genomic and epigenomic features to investigate cortical development. Together, these findings show that PromptBio can coordinate complex biomedical analyses under researcher oversight, offering an accessible approach for accelerating scientific iteration and transforming research questions into transparent, reusable computational workflows.
Related
- An agentic AI environment to support biomedical research in collaborative academic environments
- Total Integration – reviving the Biomedical Platform as an analytical tool
- Enrich.AI: An Agentic AI Platform for Interactive GeneSet Enrichment Analysis and Biological Interpretation
- NeuroPilot: Longitudinal intravital neural imaging across scales and modalities through an agentic platform
- TIAgent: From Natural Language to Grounded Agentic Workflows for Computational Pathology
BibTeX
@misc{zhang2026promptbio,
title = {{PromptBio: An Agentic Platform for End-to-End Computational Biomedical Research}},
author = {M. Zhang and W. Gu and B. Han and W. Guo and J. Chen and X. Zhou and Y. Leng and Y. Ma and K. Li and J. Zheng and H. Shishir and W. Wang and A. Huang and K. Shashidhar and X. Yang},
howpublished = {bioRxiv},
year = {2026},
month = sep,
doi = {10.64898/2026.09.02.748774},
url = {https://www.biorxiv.org/content/10.64898/2026.09.02.748774v1},
note = {CC BY-NC-ND},
}
RIS
TY - UNPB
TI - PromptBio: An Agentic Platform for End-to-End Computational Biomedical Research
AU - M. Zhang
AU - W. Gu
AU - B. Han
AU - W. Guo
AU - J. Chen
AU - X. Zhou
AU - Y. Leng
AU - Y. Ma
AU - K. Li
AU - J. Zheng
AU - H. Shishir
AU - W. Wang
AU - A. Huang
AU - K. Shashidhar
AU - X. Yang
PY - 2026
DA - 2026/09/02
PB - bioRxiv
DO - 10.64898/2026.09.02.748774
UR - https://www.biorxiv.org/content/10.64898/2026.09.02.748774v1
AB - Modern biomedical research increasingly depends on complex computational analyses, yet translating a scientific question into a reliable workflow still requires substantial technical expertise and manual coordination. PromptBio is a multi-agent AI platform available through a web portal at https://promptbio.ai that addresses this challenge. Through natural-language interaction, its agent harness, PromptGenie, translates a research objective into an inspectable plan, executes the required research and analysis, and adapts subsequent steps as evidence and results emerge. PromptBio integrates reasoning with managed execution while preserving human oversight and a traceable record of the research process. It can apply validated methods, construct custom analyses, and incorporate external workflows, enabling researchers to move beyond fixed pipelines while supporting reproducibility. We evaluate the platform through benchmarks of end-to-end bioinformatics analysis and biomedical deep research, validation of representative omics and machine-learning skills, and a hypothesis-driven regulatory-genomics case study. PromptGenie achieved higher analytical accuracy and stronger evidence retrieval and synthesis than the comparison agents, while the evaluated domain skills produced results consistent with established methods. The case study further demonstrates how PromptBio can integrate human-specific genomic and epigenomic features to investigate cortical development. Together, these findings show that PromptBio can coordinate complex biomedical analyses under researcher oversight, offering an accessible approach for accelerating scientific iteration and transforming research questions into transparent, reusable computational workflows.
ER -