Ab initio molecular dynamics simulation reveals the influence of entropy effect on Co@BEA zeolite-catalyzed dehydrogenation of ethane

The C–H bond activation in alkane dehydrogenation reactions is a key step in determining the reaction rate.To understand the impact of entropy,we performed ab initio static and molecular dynamics free energy simulations of ethane dehydrogenation over Co@BEA zeolite at different temperatures.AIMD simulations showed that a sharp decrease in free energy barrier as temperature increased.Our analysis of the temperature dependence of activation free energies uncovered an unusual entropic effect accompanying the reaction.The unique spatial structures around the Co active site at different temperatures influenced both the extent of charge transfer in the transition state and the arrangement of 3d orbital energy levels.We provided explanations consistent with the principles of thermodynamics and statistical physics.The insights gained at the atomic level have offered a fresh interpretation of the intricate long-range interplay between local chemical reactions and extensive chemical environments.

Overhead-free Noise-tolerant Federated Learning: A New Baseline

Federated learning (FL) is a promising decentralized machine learning approach that enables multiple distributed clients to train a model jointly while keeping their data private. However, in real-world scenarios, the supervised training data stored in local clients inevitably suffer from imperfect annotations, resulting in subjective, inconsistent and biased labels. These noisy labels can harm the collaborative aggregation process of FL by inducing inconsistent decision boundaries. Unfortunately, few attempts have been made towards noise-tolerant federated learning, with most of them relying on the strategy of transmitting overhead messages to assist noisy labels detection and correction, which increases the communication burden as well as privacy risks. In this paper, we propose a simple yet effective method for noise-tolerant FL based on the well-established co-training framework. Our method leverages the inherent discrepancy in the learning ability of the local and global models in FL, which can be regarded as two complementary views. By iteratively exchanging samples with their high confident predictions, the two models “teach each other” to suppress the influence of noisy labels. The proposed scheme enjoys the benefit of overhead cost-free and can serve as a robust and efficient baseline for noise-tolerant federated learning. Experimental results demonstrate that our method outperforms existing approaches, highlighting the superiority of our method.

Cell-free expression of NO synthase and P450 enzyme for the biosynthesis of an unnatural amino acid L-4-nitrotryptophan

Cell-free system has emerged as a powerful platform with a wide range of in vitro applications and recently has contributed to express metabolic pathways for biosynthesis.Here we report in vitro construction of a native biosynthetic pathway for L-4-nitrotryptophan(L-4-nitro-Trp)synthesis using an Escherichia coli-based cell-free protein synthesis(CFPS)system.Naturally,a nitric oxide(NO)synthase(TxtD)and a cytochrome P450 enzyme(TxtE)are responsible for synthesizing L-4-nitro-Trp,which serves as one substrate for the biosynthesis of a nonribosomal peptide herbicide thaxtomin A.Recombinant coexpression of TxtD and TxtE in a heterologous host like E.coli for L-4-nitro-Trp production has not been achieved so far due to the poor or insoluble expression of TxtD.Using CFPS,TxtD and TxtE were successfully expressed in vitro,enabling the formation of L-4-nitro-Trp.After optimization,the cell-free system was able to synthesize approximately 360μM L-4-nitro-Trp within 16 h.Overall,this work expands the application scope of CFPS for study and synthesis of nitro-containing compounds,which are important building blocks widely used in pharmaceuticals,agrochemicals,and industrial chemicals.