In a comprehensive review published in MedComm – Future Medicine, researchers elucidate the transformative impact of deep learning technologies on protein structure prediction. This review explores the evolution from traditional computational methods to advanced deep learning models, such as AlphaFold 3, which have markedly improved the accuracy of predicting protein structures from amino acid sequences. These developments hold significant implications for bioinformatics, enhancing drug discovery, understanding disease mechanisms, and enabling the design of novel biological systems.

In a significant advance for sustainable energy, scientists have developed a new method to convert carbon dioxide (CO₂) into formate, a valuable chemical, through a solar-driven process. This innovative approach boosts both the efficiency and selectivity of CO₂ conversion, presenting a dual solution for renewable energy utilization and reducing greenhouse gas emissions.

Metabolic associated fatty liver disease is characterized by liver fat accumulation and linked to mitochondrial dysfunction. This study shows that tetradecylthioacetic acid (TTA) significantly reduces meldonium-induced hepatic steatosis by enhancing mitochondrial function and restoring fatty acid metabolism. TTA improves delta-6 desaturase and elongase activities, indicating that plasma levels of γ-linolenic acid and estimated delta-6 desaturase and n-6 elongase indexes may serve as prognostic markers for fatty liver development.

A ray tracing model was established to calculate the transmission trajectory of laser beams in multipass cell (MPC). Types of MPCs with dense spot patterns were developed, and related sensor performance studies were conducted. A near-infrared light-induced thermoelastic spectroscopy (LITES)-based methane sensor based on a Raman fiber amplifier and a four-concentric-circle pattern MPC was built, and achieved excellent detection performance. This study is expected to bring new applications in environmental monitoring and fire warning.