Toward informed batch correction for single-cell transcriptome integration - Nature

We’re sharing this article early to provide faster access to peer-reviewed, accepted research. It is citable and carries a permanent DOI. This version is subject to further edits and will be replaced automatically by the final Version of Record. All legal disclaimers apply. AbstractErigeron breviscapus is a widely utilized medicinal herb in traditional Chinese medicine, characterized by high annual demand and notable pharmacological activities. Light intensity is a critical environmental factor influencing secondary metabolism in medicinal plants. However, the effects of light intensity on the accumulation of active ingredients in E. breviscapus remain largely unexplored. In this study, we examined the transcriptomic and metabolomic responses of E. breviscapus to five levels of light intensity (50, 100, 200, 400, 800 μmol·m−2·s−1). The results revealed that higher light intensity (EB-L800) significantly increased the accumulation of scutellarin, chlorogenic acid, isochlorogenic acid A, and isochlorogenic acid C. A total of 1,334 metabolites were detected, and 127.10 Gb of high-quality clean data were generated through transcriptomic sequencing.

(Toronto, October 16, 2026) — JMIR Publications this week released three News and Perspectives features on the latest digital health news: stories on how informed consent must be rethought for clinical trials involving AI; Canada’s proposed Safe Social Media Act and the Canadian Medical Association’s support; and a promising new autonomous oxygen titration device for health care. In "Informed Consent in the AI Age: Why and How Documentation Must Change", JMIR Correspondent and researcher Hankun Su argues that traditional informed consent frameworks, which tend to assume that risks to patient safety are static and that clinicians will be able to fill in any gaps in understanding, are inadequate for clinical trials that use AI. “Documents must anticipate questions clinicians cannot answer and provide accessible responses within the text itself,” he writes. Su and his colleagues developed an informed consent framework for AI-related clinical trials that mandates transparency and disclosure, uses tiered standards based on risk, and employs dynamic consent mechanisms—one of many possible frameworks.