Gastrointestinal (GI) cancers account for approximately one‑third of cancer‑related deaths worldwide, and outcomes remain poor despite advances in surgery, chemotherapy, radiotherapy, and immunotherapy. Gut microbiota and fecal microbiota transplantation (FMT) are increasingly recognized as potential therapeutic adjuvants. FMT may enhance antitumor efficacy and reduce treatment‑related toxicity by improving antigen presentation, reshaping the tumor microenvironment (TME), and preserving intestinal barrier function. Early clinical evidence suggests that FMT combined with immune checkpoint inhibitors (ICIs), chemotherapy, or radiotherapy may improve response rates and alleviate toxicity in selected patients. However, donor microbiota heterogeneity, unpredictable engraftment, and lack of validated predictive models hinder clinical translation. This review proposes a functional donor screening framework—the Healthy Donor‑derived Microbiota Xenograft (HDMX) model and its clinical application, XenoFMT—to advance FMT toward precision intervention.Gastrointestinal (GI) cancers account for approximately one‑third of cancer‑related deaths worldwide, and outcomes remain poor despite advances in surgery, chemotherapy, radiotherapy, and immunotherapy. Gut microbiota and fecal microbiota transplantation (FMT) are increasingly recognized as potential therapeutic adjuvants. FMT may enhance antitumor efficacy and reduce treatment‑related toxicity by improving antigen presentation, reshaping the tumor microenvironment (TME), and preserving intestinal barrier function. Early clinical evidence suggests that FMT combined with immune checkpoint inhibitors (ICIs), chemotherapy, or radiotherapy may improve response rates and alleviate toxicity in selected patients. However, donor microbiota heterogeneity, unpredictable engraftment, and lack of validated predictive models hinder clinical translation. This review proposes a functional donor screening framework—the Healthy Donor‑derived Microbiota Xenograft (HDMX) model and its clinical application, XenoFMT—to advance FMT toward precision intervention.

A Pusan National University study reveals that sustainability communication may depend not only on the message itself but also on the sequence in which the environmental information is presented. The study reveals that an initial exposure to environmental degradation can heighten early attentional engagement and emotional arousal, while a subsequent positive green solution creates a contrast-based shift in emotional valence that drives stronger pro-environmental behaviors.  

A study co-authored by Prof. Piotr Kuś from Wroclaw Medical University investigated the antifungal activity of Polish propolis and its interactions with commonly used antifungal drugs. The researchers identified 77 compounds in the propolis extract and tested it against four yeast pathogens responsible for candidiasis, including the drug-resistant Candida auris.

The most promising results were observed when propolis was combined with amphotericin B. The combination showed synergistic activity against all tested yeast species, suggesting that propolis may enhance the drug’s antifungal effect. However, not all combinations were beneficial. Some antifungal drugs showed no meaningful interaction with propolis, while others, including caspofungin, clotrimazole and ketoconazole, showed signs of antagonism.

The study also provided early evidence that ethanolic propolis extract can eliminate C. auris under laboratory conditions. The findings are limited to in vitro experiments, so propolis cannot yet be recommended as an adjunct to antifungal treatment. Further research, including extract standardization, preclinical studies and eventually clinical trials, is needed to determine whether these findings could lead to new combination therapies or drug candidates.

No-code biological modeling: BSC researchers developed open-access tools (MCP servers) that connect AI agents with specialized computational biology software, allowing complex multicellular models to be built through simple conversation.

From months to 10 minutes: The framework enables experimental biologists to generate initial 3D tumor digital twin prototypes in under 10 minutes without writing a single line of code or mastering complex software.

Accelerating cancer research: Published in npj Systems Biology and Applications, the study demonstrates how targeted, iterative AI interactions ensure reproducibility, paving the way for AI-assisted modeling of complex diseases.

A study in Geodesy and Geodynamics used satellite radar to examine the November 2025 Hayli Gubbi eruption in Ethiopia’s Afar Depression. Researchers combined NISAR’s left-looking radar with Sentinel-1’s right-looking observations to reconstruct three-dimensional ground movement. Although the eruption occurred at Hayli Gubbi, the strongest deformation was detected at Erta Ale, where the summit crater partially collapsed and the ground sank inward. Modeling linked this to two shallow dikes beneath Erta Ale, both about 1 km deep. The event also reactivated nine nearby faults. Compared with a larger eruption earlier in 2025, the November event was more localized and relied mainly on shallow magma storage.