Traditional point-sampling methods for soil elemental analysis often struggle to capture field-scale spatial variations and rely on destructive, labor-intensive bulk density measurements. Inelastic Neutron Scattering (INS) offers rapid, non-destructive, in situ measurements of soil elemental composition over large volumes. A unique advantage of this neutron-based method is the potential to measure all major soil components simultaneously, allowing for the simultaneous estimation of bulk density and water content alongside elemental carbon. While standard INS yields bulk concentrations over a given volume, spatially resolved measurements require techniques such as Associated Particle Imaging (API). INS-API techniques provide non-destructive access to depth-resolved information, but the quantitative accuracy of this technique depends heavily on correcting for signal attenuation. While gamma-ray attenuation is readily predicted from soil density, neutron attenuation is complex and highly sensitive to hydrogen content, confounding carbon measurements in soils. In this study, we use Monte Carlo simulations of soils with varied compositions, bulk densities, and water contents to model neutron attenuation and develop a simple predictive model requiring only dry bulk density and volumetric water content. We find that such a simplification achieves accuracy within 10% at 30cm depth for simulated soils, and validate the model experimentally to 18cm depth using an INS–API system with controlled soil columns. This approach enables practical correction of INS–API measurements, laying the groundwork for a self-consistent framework to monitor soil carbon stocks independent of moisture fluctuations.
We develop an effective field theory characterizing the impact of decoherence on states with abelian topological order and on their capacity to protect quantum information. The decoherence appears as a temporal defect in the double topological quantum field theory that describes the pure density matrix of the uncorrupted state, and it drives a boundary phase transition involving anyon condensation at a critical coupling strength. The ensuing decoherence-induced phases and the loss of quantum information are classified by the Lagrangian subgroups of the double topological order. Our framework generalizes the error recovery transitions, previously derived for certain stabilizer codes, to generic topologically ordered states and shows that they stem from phase transitions in the intrinsic topological order characterizing the mixed state.
The growing use of ionizing radiation across medical and industrial fields has intensified the demand for lead-free, flexible, and sustainable radiation shielding materials. However, current development strategies remain heavily reliant on empirical trial-and-error, which is both time-consuming and resource-intensive. Here, we report a machine learning-assisted Monte Carlo simulation strategy that enables rapid and accurate optimization of metal filler compositions for efficient X-ray attenuation across a broad energy range (40–120 kV). Guided by this artificial intelligence (AI)-driven approach, we developed polyvinyl alcohol (PVA)-based gels containing uniformly dispersed Bi/W/Gd2O3 nanoparticles, which form within 1 min at −20°C using a PVA-DMSO/H2O co-solvent system. The optimized gel with 50 wt
Pathogenic DEGS1 variants have been reported in individuals with autosomal recessive hypomyelinating leukodystrophy 18 (HLD18; MIM# 618404). Here we describe three participants with HLD features and a previously unreported homozygous DEGS1 5′ splice site variant, c.825+4_825 + 5delAGinsTT (NM_003676.4). We used next-generation DNA and transcriptome sequencing, cell-based splicing assays, and tandem mass spectrometry to detect and characterize the variant’s impact on DEGS1 expression. We then performed RNA structure probing and conventional antisense oligonucleotide screening to investigate molecular mechanisms for potential therapeutic intervention. We show that the splice site variant: (1) was sufficient to induce exon two skipping in most detected transcripts; (2) resulted in structural changes to the 5′ and 3′ splice site regions using RNA structure probing; and (3) corresponds to plasma sphingolipid profiles consistent with loss of sphingolipid delta(4)-desaturase activity. Our RNA and lipidomic evidence proved that the DEGS1 variant c.825+4_825 + 5delAGinsTT is pathogenic and suggested a mechanistic model that explains how exon two skipping is induced.
The tens of millions of spectra being captured by the Dark Energy Spectroscopic Instrument (DESI) provide tremendous discovery potential. In this work we show how Machine Learning, in particular Variational Autoencoders (VAE), can detect anomalies in a sample of approximately 200,000 DESI spectra comprising galaxies, quasars and stars. We demonstrate that the VAE can compress the dimensionality of a spectrum by a factor of 100, while still retaining enough information to accurately reconstruct spectral features. We then detect anomalous spectra as those with high reconstruction error and those which are isolated in the VAE latent representation. The anomalies identified fall into two categories: spectra with artefacts and spectra with unique physical features. Awareness of the former can help to improve the DESI spectroscopic pipeline; whilst the latter can lead to the identification of new and unusual objects. To further curate the list of outliers, we use the Astronomaly package which employs Active Learning to provide personalised outlier recommendations for visual inspection. In this work we also explore the VAE latent space, finding that different object classes and subclasses are separated despite being unlabelled. We demonstrate the interpretability of this latent space by identifying tracks within it that correspond to various spectral characteristics. For example, we find tracks that correspond to increasing star formation and increase in broad emission lines along the Balmer series. In upcoming work we hope to apply the methods presented here to search for both systematics and astrophysically interesting objects in much larger datasets of DESI spectra.
