元数据
微生物群
审计
序列(生物学)
注释
万维网
数据科学
计算机科学
疾病
人类疾病
医学
人体微生物群
审计跟踪
寄主(生物学)
生物信息学
肠道微生物群
梅德林
计算生物学
生物
全基因组测序
情报检索
作者
Samuel Degregori,Isabella Huang,Lynn H. Fetcinko,Harrison Martel,Dmitry Kisselev,Harrison Gu,David G. Kobobel,Emily Song,Ariadne Garcia Reyes,Malleeka T. Suy,Khanh Nguyen,Ariana J Hampton,Jack Anderson,Mohak Prakash,Jeffrey Bohrer,Nathália Maria Afonso Silveira FRANCO,Ritvik Vudatha,Cynthia Sofía,Weili Lu,Nicolás A. Zepeda
出处
期刊:iMeta
[Wiley]
日期:2026-11-18
摘要
A systematic, PRISMA-guided review led by our consortium screened 23,781 human microbiome papers and procured 2896 studies published from 2012 to 2025 that were then scored by reviewers into four Biological Host-Associated AI-Ready Metadata (BHAM) tiers. Only 5.9% of studies deposited sequence data with per-sample host disease status, age, and sex, and 63.1% shared no reusable sequence data, so that 85.5% of roughly 720,000 samples cannot be reused for cross-cohort or artificial intelligence (AI) analyses without contacting the authors. Sequence deposition improved modestly over time, whereas per-sample annotation did not. Journals, repositories, and researchers should require structured, machine-readable metadata at submission, supported by AI-assisted curation.
科研通智能强力驱动
Strongly Powered by AbleSci AI