The machine learning-derived model was 88% accurate in ranking preoperative patients according to risk for postoperative acute kidney injury.

The artificial intelligence model Enformer was retrained with tissue-specific datasets; this "transfer learning" approach outperformed the base model in identifying disease-associated genes.

These findings shed light on disruptions in normal blood cell development (hematopoiesis) that may be shared by sickle cell disease and clonal hematopoiesis.

One immune cell signaling molecule was increased in active versus inactive Crohn’s disease in both endoscopic and histologic activity assessments, suggesting it could serve as a potential noninvasive marker of disease activity.

Study findings suggest a new therapeutic approach for a rare digestive disorder, microvillus inclusion disease.

Kindlins regulate cell signaling during the formation of ureteric buds, a key step in the development of the kidneys’ urine collection system.

A novel AI framework found documented intolerance to statins in 6.4% of all adult patients, contraindications in 0.7%, and statin deferral in 2.9%.