The University of Maryland, Baltimore (UMB) has recommended the retraction of four papers on maternal diabetes and neural tube formation, according to reporting by The Transmitter published on 16 October 2026 and follow-up coverage in The Scientist. The notices cite image duplication and other data integrity problems. The case is a current example of how an institutional investigation, a publisher and post-publication reviewers interact when published figures are questioned.
What happened
The four retracted papers appeared in Communications Biology (2022), Biochemical and Biophysical Research Communications (2015), Reproductive Toxicology (2019) and Cell Death and Disease (2020). Peixin Yang, a UMB professor of obstetrics, gynecology and reproductive sciences, is the corresponding author on all four, according to The Transmitter and The Scientist. E. Albert Reece and Wei-Bin Shen, also at UMB, co-authored some of them.
The Communications Biology notice, which is publicly available at PubMed Central, retracts “Functional cargos of exosomes derived from Flk-1+ vascular progenitors enable neurulation and ameliorate embryonic anomalies in diabetic pregnancy” (published 1 July 2022). The notice states that UMB recommended the retraction. It says a peer-review investigation committee, overseen by UMB’s Research Integrity Office, found that the data in Figure 5d were mislabelled and came from experiments different from those the article described, and that not all of the relevant original data was available for inspection, so the committee could not confirm the data’s integrity. The editors state that they no longer have confidence in the results and conclusions. The notice also says the authors have not explicitly stated whether they agree or disagree with the retraction. PubMed Central lists the notice’s citation date as 6 August 2025.
The Transmitter reports that the notices for the other three papers cite image duplication and other data integrity issues, and that the notices for the 2019, 2020 and 2022 papers do not state that Yang committed research misconduct, as described by his attorney. The Transmitter also reports that UMB denied its public-records request for the investigation report, citing an “ongoing personnel matter”.
Context
The Transmitter reported in June 2025 that data sleuths had flagged 25 papers from the group for possible image duplication or manipulation. Reviewers have since flagged a total of 44 studies on PubPeer, and Yang is the only author common to all 44, according to The Transmitter. Flags on PubPeer are post-publication comments, not findings.
The Transmitter also reports that Yang’s attorney, Alexa Millinger of Hinckley Allen, sent letters dated 3 and 18 September 2026 demanding that its earlier article be corrected or retracted, saying it creates a false impression of misconduct. The Transmitter’s article describes the dispute, and the notices themselves do not state a misconduct finding against any individual.
What it means for researchers
- Keep original data retrievable. The Communications Biology notice turns on the availability of original data for inspection. Raw images, lab notebook entries and instrument files that can be matched to published figures are what allow a committee to confirm or resolve a concern.
- Co-authors share exposure. A retraction attaches to every author of the paper, whatever their role in the questioned figure. Reviewing key figures against source data before submission is a practical safeguard.
- Retraction is not the same as a misconduct finding. A retraction records that the editors no longer trust the results. A formal research misconduct determination is a separate process with its own standards.
What it means for universities and research offices
- Institutional recommendations drive retractions. Here the publisher acted at the university’s recommendation after an internal investigation. Research integrity officers (see research integrity officer) should expect journals to ask for the committee’s conclusions and should plan how those are communicated.
- Confidentiality and transparency pull against each other. UMB cited a personnel matter in declining to release its report, while the public notices carry only summary findings. Offices need a clear policy on what is disclosed to journals, funders and the press while a review is open.
- Portfolio-wide questions. With dozens of papers flagged, offices must decide whether to review a group’s wider output and how to track related editorial actions.
- Data retention policies matter. Missing original data can by itself prevent a committee from confirming integrity, so retention periods and storage standards deserve review. See also image duplication for how such concerns are typically described.
Sources
- The Transmitter, “University probe leads to retraction of four neural tube studies” (16 October 2026), read 12 November 2026.
- The Scientist, “Four neurodevelopment papers retracted over data integrity issues”, read 12 November 2026.
- Retraction Note, Communications Biology 8:1162 (2025), PubMed Central, read 12 November 2026.
- The Transmitter, “More than two dozen papers by neural tube researcher come under scrutiny” (June 2025), read 12 November 2026.
