Research Briefing

Different cells can have very different three-dimensional morphologies. We present the computational framework u-signal3D that calculates the spatial scales at which molecules are organized on the surfaces of heterogeneously shaped cells, enabling high-throughput analyses and subsequent machine learning applications.

Graph neural networks (GNNs) present a promising route for machine learning of solid-state materials’ properties, but methods capable of directly predicting defect properties from ideal, defect-free structures are needed. A GNN developed for direct defect property predictions enables high-throughput screening of redox-active oxides for energy applications and beyond.

The dissipation and bending of light waves by atmospheric turbulence adversely affects infrared imaging, leading to grayscale drift, distortion, and blurring. A deep learning method has been developed to both extract the two-dimensional atmospheric turbulence strength fields and obtain clear and stable images from turbulence-distorted infrared images.

Increasing the number of parameters in a quantum neural network leads to a computational ‘phase transition’, beyond which training the network becomes significantly easier. An algebraic theory has been developed for this overparametrization phenomenon and predicts its onset above a certain parameter threshold.

GRAPE is a software resource for graph processing, learning and embedding that is orders of magnitude faster than existing state-of-the-art libraries. GRAPE can quickly process real-world graphs with millions of nodes and billions of edges, enabling complex graph analyses and research in graph-based machine learning and in diverse disciplines.

By conducting single-cell meta-analyses of inflammatory bowel disease, we identify rare or less-characterized cell subtypes linked to GWAS risk genes and therapeutic targets and dissect the commonalities and differences between ulcerative colitis and Crohn’s disease. Consequently, we present an interactive and user-friendly platform for the research community.

We often encounter mental conflict in our lives. Such mental conflict has long been regarded as subjective. However, a machine learning method can be used to quantify the temporal dynamics of conflict between reward and curiosity from behavioral time-series.

A graph neural network — GAME-Net — has been developed to predict the adsorption energy of organic molecules on metal surfaces, which is a key descriptor of heterogeneous catalytic activity. This method allows for the study of large molecules derived from raw materials such as plastic waste, avoiding the use of costly and time-intensive first-principles simulations.

A rotational and time-reversal equivariant neural network designed to represent the spin–orbital density functional theory Hamiltonian as a function of the atomic and magnetic structure enables ab initio electronic-structure calculations of magnetic superstructures. These calculations can efficiently and accurately predict subtle magnetic effects in various chemical environments.

A machine learning algorithm has been developed to capture and analyze rare molecular processes, revealing how molecules self-organize and function. The algorithm is general and can be applied whenever a dynamic system has a notion of ‘likely fate’.

This work involved the design of a multi-view manifold learning algorithm that capitalizes on various types of structure in high-dimensional time-series data to model dynamic signals in low dimensions. The resulting embeddings of human functional brain imaging data unveil trajectories through brain states that predict cognitive processing during diverse experimental tasks.

We present a computational method to generate a single-cell-resolution model of human brain regions starting from microscopy images. The developed method has been benchmarked to reconstruct the CA1 region of a right human hippocampus, including anatomical cell organization, connectivity, and network activity.

We propose a minimal and analytically tractable class of neural networks, the adaptive Ising class. By inferring the model’s parameters from resting-state brain activity recordings, we show that scale-specific oscillations and scale-free avalanches can coexist in resting brains close to a non-equilibrium critical point at the onset of self-sustained oscillations.

Determining whether a drug candidate has sufficient affinity to its target is a critical part of drug development. A purely physics-based computational method was developed that uses non-equilibrium statistical mechanics approaches alongside molecular dynamics simulations. This technique could enable researchers to accurately estimate the binding affinities of potential drug candidates.

We used computational models built using neural networks to predict what brain areas process the new meaning that emerges when words are combined. The brain activity evoked by this composed meaning was detected only with some brain recording modalities, a finding that might have consequences for brain–computer interfaces.

A universal interatomic potential for the periodic table has been developed by combining graph neural networks with three-body interactions. This M3GNet potential can perform structural relaxations, dynamic simulations and property predictions for materials across a diverse chemical space.

The surface energy cannot be assigned to each direction in low-symmetry crystals, making it impossible to predict their shapes by any known methods. Now, combining incomputable energies in an algebraic system, complemented by closure equations, it is possible to predict the equilibrium shape of any crystal.

We developed a computational method to reveal the drug-induced single-cell transcriptomic landscape. This algorithm enabled us to impute unknown drug-induced single-cell gene expression profiles using tensor imputation, predict cell type-specific drug efficacy, detect cell-type-specific marker genes, and identify the trajectories of regulated biological pathways while considering intercellular heterogeneity.

In social interactions, individuals are often tempted to ‘free ride’ (benefit without paying) on other group members’ contributions. A new mathematical framework suggests a strategy based on cumulative reciprocity that can help to sustain mutual cooperation.

Designing efficient bike path networks requires balancing multiple opposing constraints such as cost and safety. An adaptive demand-driven inverse percolation approach is proposed to generate efficient network structures by explicitly taking into account the demands of cyclists and their route choice behavior based on safety preferences.