Browse Articles
Discover research articles across all indexed journals
In-silico comparison of a diffusion model with conventionally trained deep networks for translating 64mT to 3T brain FLAIR
Spatial joint profiling of DNA methylome and transcriptome in tissues
Abstract The spatial resolution of omics analyses is fundamental to understanding tissue biology 1–3 . The capacity to spatially profile DNA methylation, which is a canonical epigenetic mark extensively implicated in transcriptional regulation 4,5 , is lacking. Here we introduce a method for whole-genome spatial co-profiling of DNA methylation and the transcriptome of the same tissue section at near single-cell resolution. Applying this technology to mouse embryogenesis and the postnatal mouse brain resulted in rich DNA–RNA bimodal tissue maps. These maps revealed the spatial context of known methylation biology and its interplay with gene expression. The concordance and distinction in spatial patterns of the two modalities highlighted a synergistic molecular definition of cell identity in spatial programming of mammalian development and brain function. By integrating spatial maps of mouse embryos at two different developmental stages, we reconstructed the dynamics that underlie mammalian embryogenesis for both the epigenome and transcriptome, revealing details of sequence-, cell-type- and region-specific methylation-mediated transcriptional regulation. This method extends the scope of spatial omics to include DNA cytosine methylation, enabling a more comprehensive understanding of tissue biology across development and disease.
Design of direction-independent hydrovoltaic electricity generator based on all-foam asymmetric electrode
The role of ROSE-enhanced ERCP-guided biopsy in diagnosing biliary stricture
Long-read RNA-seq demarcates cis- and trans-directed alternative RNA splicing
A computer vision framework for proactive anomaly detection and risk reduction in airport baggage logistics
Hippocampus supports multi-task reinforcement learning under partial observability
Abstract Mastering navigation in environments with limited visibility is crucial for survival. Although the hippocampus has been associated with goal-oriented navigation, its role in real-world behaviour remains unclear. To investigate this, we combined deep reinforcement learning (RL) modelling with behavioural and neural data analysis. First, we trained RL agents in partially observable environments using egocentric and allocentric tasks. We show that agents equipped with recurrent hippocampal circuitry, but not purely feedforward networks, learned the tasks in line with animal behaviour. Next, we used dimensionality reduction of the agents’ internal representations to extract components reflecting reward, strategy, and temporal representations, which we validated experimentally against hippocampal recordings from rats. Moreover, hippocampal RL agents predicted state-specific trajectories, mirroring empirical findings. In contrast, agents trained in fully observable environments failed to capture experimental observations. Finally, we show that hippocampal-like RL agents demonstrated improved generalisation across novel task conditions. In summary, our findings suggest an important role of hippocampal networks in facilitating reinforcement learning in naturalistic environments.
Combining remote sensing with local knowledge is vital for understanding forest change in West Africa
Abstract Understanding tropical forest change requires integrating satellite observations with insights from forest-dependent communities. In West Africa, where deforestation and degradation unfold within complex social-ecological systems, conventional monitoring often overlooks community insights. We combined Landsat-derived forest cover data (2000–2022) with household surveys from 2,621 respondents across nine forest patches, applying a convergence matrix to compare satellite trajectories with local knowledge of forest change. Two themes were analyzed: forest cover loss and forest regrowth. Sites were classified as full convergence, partial convergence, or dissonance based on directional agreement and the proportion of community responses. Full convergence occurred in four sites where > 65% of respondents reported forest loss consistent with satellite-detected declines (− 2.8% to − 13.9% cover). Partial convergence characterized the two sites with mixed local responses and modest net satellite-image changes (< 5%). Dissonance emerged in three sites where satellite-detected stability (< 2% net change, p ≥ 0.10) contrasted with > 65% of respondents reporting degradation. Across all sites, satellite-detected regrowth was minimal, though some communities described localized recovery. These findings show that convergence is strongest for deforestation, while divergences are concentrated around degradation, underscoring the diagnostic value of local knowledge and the limitations of medium-resolution imagery. Embedding such knowledge in monitoring frameworks is therefore both an ethical imperative and a strategic necessity for adaptive forest governance.