Browse Articles
Discover research articles across all indexed journals
Sharing benefits of research is key to effective science communication
Targeting FOXM1 condensates reduces breast tumour growth and metastasis
Rational design of phytovirucide inhibiting nucleocapsid protein aggregation in tomato spotted wilt virus
PhD scientists who forged a career off the beaten track
An infrared-transparent textile with high drawing processed Nylon 6 nanofibers
Capturing ultrafast molecular motions and lattice dynamics in spin crossover film using femtosecond diffraction methods
Abstract A comprehensive insight into ultrafast dynamics of photo-switchable materials is desired for efficient control of material properties through light excitation. Here, we study a polycrystalline spin crossover thin film as a prototypical example and reveal the sequential photo-switching dynamics, from local molecular rearrangement to global lattice deformation. On the earliest femtosecond timescale, the local molecular structural rearrangement occurs within a constant unit-cell volume through a two-step process, involving initial Fe−ligand bond elongation followed by ligand rotation. The highly-oriented structure of the nanocrystalline films and the experimental geometry enables resolving the full anisotropic lattice structural dynamics in and out of the sample plane separately. While both molecular switching and lattice heating influence lattice volume, they exert varying degrees of impact at disparate time scales following photoexcitation. This study highlights the opportunities provided by Mega-electron-volt electron and X-ray free electron laser to advance the understanding of ultrafast dynamics of photo-switchable materials.
We use 30 billion tonnes of concrete each year — here’s how to make it sustainable
Interpreting mammalian synonymous site conservation in light of the unwanted transcript hypothesis
Abstract Mammalian genomes are biased towards GC bases at third codon positions, likely due to a GC-biased ancestral genome and the selectively neutral recombination-related process of GC-biased gene conversion. The unwanted transcript hypothesis posits that this high GC content at synonymous sites may be beneficial for protecting against spurious transcripts, particularly in species with low effective population sizes. Utilising a 240 placental mammal genome alignment and single-base resolution conservation scores, we interpret sequence conservation at mammalian four-fold degenerate sites in this context and find evidence in support of the unwanted transcript hypothesis, including a strong GC bias, high conservation at sites relating to exon splicing, less human genetic variation at conserved four-fold degenerate sites, and conservation of sites important for epigenetic regulation of developmental genes. Additionally, we show that high conservation of four-fold degenerate sites in essential developmental genes, including homeobox genes, likely relates to the low mutation rates experienced by these genes.
China’s low-altitude economy needs to grow with more caution
IL-33-activated ILC2s induce tertiary lymphoid structures in pancreatic cancer
Operando impedance spectroscopy with combined dynamic measurements and overvoltage analysis in lithium metal batteries
Synthesis of high quality two dimensional covalent organic frameworks through a self-sacrificing guest strategy
Advancing nonadiabatic molecular dynamics simulations in solids with E(3) equivariant deep neural hamiltonians
Abstract Non-adiabatic molecular dynamics (NAMD) simulations have become an indispensable tool for investigating excited-state dynamics in solids. In this work, we propose a general framework, N2AMD (Neural-Network Non-Adiabatic Molecular Dynamics), which employs an E(3)-equivariant deep neural Hamiltonian to boost the accuracy and efficiency of NAMD simulations. Distinct from conventional machine learning methods that predict key quantities in NAMD, N2AMD computes these quantities directly with a deep neural Hamiltonian, ensuring excellent accuracy, efficiency, and consistency. N2AMD not only achieves impressive efficiency in performing NAMD simulations at the hybrid functional level within the framework of the classical path approximation (CPA), but also demonstrates great potential in predicting non-adiabatic coupling vectors and suggests a method to go beyond CPA. Furthermore, N2AMD demonstrates excellent generalizability and enables seamless integration with advanced NAMD techniques and infrastructures. Taking several extensively investigated semiconductors as the prototypical system, we successfully simulate carrier recombination in both pristine and defective systems at large scales where conventional NAMD often significantly underestimates or even qualitatively incorrectly predicts lifetimes. This framework offers a reliable and efficient approach for conducting accurate NAMD simulations across various condensed materials.
How scientists can drive climate action: celebrate nature and promote hope
Rapid and accurate prediction of protein homo-oligomer symmetry using Seq2Symm
Abstract The majority of proteins must form higher-order assemblies to perform their biological functions, yet few machine learning models can accurately and rapidly predict the symmetry of assemblies involving multiple copies of the same protein chain. Here, we address this gap by finetuning several classes of protein foundation models, to predict homo-oligomer symmetry. Our best model named Seq2Symm, which utilizes ESM2, outperforms existing template-based and deep learning methods achieving an average AUC-PR of 0.47, 0.44 and 0.49 across homo-oligomer symmetries on three held-out test sets compared to 0.24, 0.24 and 0.25 with template-based search. Seq2Symm uses a single sequence as input and can predict at the rate of ~80,000 proteins/hour. We apply this method to 5 proteomes and ~3.5 million unlabeled protein sequences, showing its promise to be used in conjunction with downstream computationally intensive all-atom structure generation methods such as RoseTTAFold2 and AlphaFold2-multimer. Code, datasets, model are available at: https://github.com/microsoft/seq2symm .