Shift in key cosmic inflation measurement could be a statistical artefact
Peer-Reviewed Publication
Updates every hour. Last Updated: 24-Apr-2026 20:16 ET (25-Apr-2026 00:16 GMT/UTC)
A new Maths study from the University of Bath in the UK finds that adopting a neutral stance – such as abstaining in a vote – can speed up and stabilise group decision-making. By reducing the pool of active decision-makers, neutrality helps new consensus positions emerge faster.
A multinational research team led by researchers at Institute of Science Tokyo, RIKEN, and the University of Toronto has revealed how a tryptophan-rich allosteric communication network regulates receptor dynamics and activation of the human adenosine A2A receptor (A2AR), a major G protein-coupled receptor (GPCR) drug target. By integrating experimental functional assays and residue-specific NMR with molecular simulations and fast allostery-prediction algorithms based on rigidity theory, the team mapped long-range allosteric communication pathways linking the ligand-binding pocket to the intracellular G protein–coupling machinery and identified a central role for tryptophan residues along these pathways. The study also clarifies the functional role of the receptor’s conserved sodium-binding pocket, showing that sodium egress strongly promotes activation-related conformational states, including a precoupled state that likely prepares the receptor for productive G protein interaction. These findings deepen our understanding of GPCR activation and allostery, and may support future development of allosteric GPCR drugs.
Beyond the specific mechanism, this work addresses a major bottleneck for AI in structural biology: recent advances such as AlphaFold have transformed prediction of static protein structures, but AI still cannot reliably predict the dynamics and allosteric communication that determine function, signaling, and drug response. To help close this gap, the researchers developed and applied fast computational methods for probing allosteric and dynamic regulation in protein structures and anchored these predictions with experimental NMR validation. The resulting experimentally validated, computationally generated data on allostery and dynamics—and a scalable approach to extend these datasets across diverse receptors and conditions—provide scarce, high-value training and benchmarking data for next-generation AI models aimed at predicting protein function beyond static structure, accelerating future AI-driven prediction of protein function and the design of selective GPCR therapeutics.
Quantum computers outperform typical computers in many tasks, yet in many other tasks, classical computers have the upper hand. Researchers at Saarland University, together with industry partners BMW, Infineon and the quantum computing start-up planqc, want to combine the strengths of both types of computing. The team plans to employ a quantum computer to help classical computers handle highly complex optimization challenges from industrial practice. The project is funded by a €2.3 million grant from the Federal Ministry of Research, Technology and Space.