--- title: 'Research Daily Summary: What Happened on October 9, 2026' headline: Placental model tracks cadmium exposure and tissue dysfunction description: A review found code-sharing statements in 12.2% of clinical prediction model papers and uneven reproducibility features across repositories. url: https://thebench.bio/news/research/2026-10-09 topic: research date: '2026-10-09' covers: - '2026-10-09' date_published: '2026-10-09T12:02:58Z' date_modified: '2026-10-09T12:02:58Z' sources: 2 publisher: The Bench (thebench.bio) --- # Research Daily Summary: What Happened on October 9, 2026 **Placental model tracks cadmium exposure and tissue dysfunction** A review found code-sharing statements in 12.2% of clinical prediction model papers and uneven reproducibility features across repositories. ## Today in brief A microfluidic system built with human placental cells modeled barrier responses and tissue impairment after cadmium exposure. [1] Researchers also identified transporter proteins that may protect against toxicity and tested the system’s potential for finding metabolic biomarkers. [1] Among 3,967 clinical prediction model papers, 482 included code-sharing statements. Repository documentation and reproducibility features varied considerably. [2] ## A placental model probes cadmium’s effects Researchers used primary human placental cells in a microfluidic device to model the maternal-fetal interface. [1] With cadmium as a model toxicant, the system modeled placental barrier responses and impaired tissue function. [1] Trophoblast efflux transporters, which move substances out of cells, may help protect against cadmium-related toxicity. [1] The researchers demonstrated the system’s potential for identifying metabolic biomarkers of cadmium-induced placental dysfunction. [1] ## Code-sharing remains uncommon in prediction-model papers Of 3,967 papers citing TRIPOD or TRIPOD+AI, 482 included statements about sharing code, a review found. [2] The review assessed 380 code repositories and found wide variation in documentation and tools for reproducing analyses. [2] A README file appeared in 80.5% of repositories, while 37.6% specified software dependencies and 3.9% included tests. [2] The authors are developing TRIPOD-Code guidance on code availability and computational reproducibility. [2] ## The day in numbers - **12.2%**: Share of 3,967 eligible prediction-model papers with code-sharing statements [2] - **3.9%**: Share of assessed repositories that included tests [2] ## Also published - [How Routine Lab Markers Can Stand In for Missing Functional Status Data](), Applied Clinical Trials - [The journal at a glance: Q3 2026](), BioTechniques - [What It Takes to Harmonize AI Across Fragmented Oncology Health Systems](), Applied Clinical Trials ## Sources 1. [A bioengineered model of human placental exposure to environmental metals during pregnancy](), Nature Biomedical Engineering, 2026-10-09 2. [AI-assisted scoping review of code sharing in clinical prediction model research](), Nature Medicine, 2026-10-09 --- The Bench wrote this from the reporting cited above, with a language model drafting and editing. [1]: [2]: