🧬 Spatiotemporal transcriptome atlas of human embryos after gastrulation

Female, Humans, Male, Alleles, Brain, Embryo, Mammalian, Gastrulation, Gene Expression Profiling, Gene Expression Regulation, Developmental, Organ Specificity, Organogenesis, Single-Cell Gene Expression Analysis, Spatial Transcriptomics, Transcriptome

Researchers built a detailed spatial map of gene activity in human embryos during early organ formation. They used Stereo-seq on 77 tissue sections from 13 embryos at Carnegie stages 12–23, combining it with single-nucleus RNA sequencing to link gene expression to specific cell types. This revealed how cellular diversity drives organ-specific development, covering 50 organs and 198 substructures, and highlighted potential regulators of tissue identity. The atlas uncovered previously unknown gene roles in heart and brain development, pinpointed organs particularly vulnerable to genetic disorders, and traced how gene variants are expressed across developmental stages. This resource offers the most comprehensive view yet of the molecular choreography behind human organogenesis, providing a foundation for understanding developmental biology and disease origins.

Read the original article →

🧩 Efficient and accurate neural-field reconstruction using resistive memory

Intelligent Systems, Multilayer Perceptrons, Neural Networks, Computer, Normal Distribution, Software, Brain, Humans, Animals

Researchers tackled the problem of reconstructing complex signals (e.g., 3D medical scans, views of dynamic scenes) from sparse, incomplete data, which is key for medical imaging and AR/VR but often too slow or energy-hungry for conventional computer hardware. They built a combined software-hardware system: on the software side, they used neural fields—networks that implicitly represent signals—and shrank them via low-rank decomposition and pruning. On the hardware side, they designed a chip based on resistive memory that computes directly in memory, avoiding the usual bottleneck of moving data between memory and processor. A Gaussian encoder exploits the natural randomness of resistive memory for efficient encoding, while a quantization circuit ensures accurate weight mapping. Tests on a 40-nm prototype showed large gains in energy efficiency (up to ~32x) and parallelism (up to ~39x) across three reconstruction tasks, with no quality loss, pointing to more efficient AI-driven medical and 3D vision systems.

Read the original article →

🧬 Developmental organization of sensory and sympathetic ganglia

Animals, Neural Crest, Ganglia, Sympathetic, Cell Lineage, Mice, Humans, Ganglia, Sensory, Neural Tube, Female, Fibroblast Growth Factors, Male, Quail

Researchers traced how neural crest cells—embryonic progenitors that form diverse tissues—commit to becoming either sensory or sympathetic neurons. Using CRISPR barcoding in mice and mosaic variant analysis in humans, they found that early progenitor cells spread broadly along the body's head-to-tail axis on both sides, yet rarely produced both sensory and sympathetic cell types. Computational modeling of the human data suggested that most neural crest cells are already fate-restricted before they migrate out of the neural tube. Live imaging in quail embryos revealed that fibroblast growth factor drives this rostrocaudal dispersal across multiple axial levels. The findings support a model where fate bias largely emerges inside the neural tube, with only a small minority of migrating progenitors retaining the ability to generate both lineages. This clarifies longstanding questions about when neural crest cells commit to specific fates during development.

Read the original article →

🐟 Chromosomal fusions trigger rediploidization of autopolyploid genomes

Animals, Alleles, Carps, Chromosomes, Evolution, Molecular, Gene Duplication, Genome, Haplotypes, Meiosis, Phylogeny, Polyploidy

This study investigates how autopolyploid genomes begin the process of rediploidization, using snow carps (Schizothoracine fish), which underwent recent whole-genome duplications. By generating haplotype-resolved genomes for two divergent species and analyzing all snow carp genera, the authors show these fish share a single ancestral autotetraploidy event, with current ploidies ranging from tetraploid to icosaploid. Through comparative genomics, meiotic pairing analysis, and allele composition, they find that unbalanced chromosome fusions drove the shift from tetrasomic to disomic inheritance, creating regions with diploidized gene pairs while unfused chromosomes stayed tetraploid. Rediploidization starts at fusion sites and spreads toward chromosome arms, remaining incomplete after speciation. This produces a mix of ancestral and lineage-specific divergence of duplicated genes on syntenic chromosomes. The findings reveal a concrete chromosomal mechanism initiating rediploidization, offering insight into how duplicated genes from whole-genome duplications may diverge functionally, relevant for understanding vertebrate evolution and high-altitude adaptation.

Read the original article →

🧠 Active dissociation of intracortical spiking and high gamma activity

Animals, Neurons, Gamma Rhythm, Macaca mulatta, Action Potentials, Brain-Computer Interfaces, Male, Synaptic Potentials, Models, Neurological

This study tested two competing explanations for the origin of high gamma-band activity (HGA) recorded from the cortex: whether it reflects summed local neuronal spikes near the electrode or summed postsynaptic potentials from larger neuronal populations. Monkeys were trained to voluntarily separate local spiking from HGA on a single electrode using a brain-machine interface. They could successfully decouple the two signals, which contradicts the idea that HGA is simply a byproduct of nearby spiking. Instead, HGA was found to track correlated firing among neurons distributed across several millimeters of cortex, with the neurons most involved in this coordinated activity contributing most to spike-triggered HGA and doing so before that signal. The authors conclude that HGA primarily reflects postsynaptic potentials generated by synchronized activity from a broad network of neurons rather than local spikes.

Read the original article →

🧠 Plasticity and language in the anaesthetized human hippocampus

Humans, Hippocampus, Language, Neuronal Plasticity, Male, Female, Neurons, Adult, Semantics, Consciousness, Unconsciousness, Local Field Potential Measurement, Models, Neurological, Anesthesia, General, Acoustic Stimulation, Young Adult, Discrimination, Psychological

The researchers probed whether complex pattern recognition in the brain can operate without consciousness. They recorded neural activity in the hippocampus of anaesthetized patients using high-density Neuropixels electrodes while playing tones and language stimuli. Despite patients being under general anaesthesia, hippocampal neurons and local oscillations still detected oddball tones against a regular series, and this neural response to the deviant tones strengthened over about ten minutes, indicating plasticity. A computational recurrent neural network model suggested that such oddball discrimination emerges naturally from flexible tone learning. When language was played, the same recordings encoded semantic and grammatical features of the speech, even anticipating upcoming words' meanings. The results indicate that the hippocampus, though far from primary sensory areas, processes complex sensory information—including semantic prediction—without conscious awareness, challenging assumptions that higher-order pattern recognition requires consciousness and informing how deep the unconscious brain processes stimuli.

Read the original article →

🦠 Lethal plague outbreaks in Lake Baikal hunter-gatherers 5,500 years ago

Child, Female, Humans, Male, Disease Outbreaks, History, Ancient, Lakes, Pedigree, Phylogeny, Plague, Siberia, Yersinia pestis, Yersinia pseudotuberculosis

Researchers identified the plague bacterium Yersinia pestis in ancient human remains from four hunter-gatherer cemeteries near Lake Baikal in Siberia, dating back to about 5,500 years ago. They detected the infection in 39% of the sampled individuals, with evidence that small family groups were hit, suggesting person-to-person transmission and outbreaks occurring within a single generation. The disease caused severe acute illness, especially in children aged 8 to 11. These ancient strains lacked the later bubonic plague adaptations and showed genetic differences, including in the ypm superantigen locus. Their analysis places these strains as ancestral to known Y. pestis and pushes back the bacterium's emergence to at least 5,700 years ago. This shows lethal plague outbreaks happened much earlier than assumed and challenges the idea that dense agricultural populations were required for plague epidemics, since these outbreaks occurred among scattered hunter-gatherer groups.

Read the original article →

🧬 Expanding the human proteome with microproteins and peptideins

Humans, Evolution, Molecular, Micropeptides, Molecular Sequence Annotation, Open Reading Frames, Peptides, Protein Biosynthesis, Proteome, Proteomics

The TransCODE Consortium mapped which non-canonical open reading frames (ncORFs) actually produce proteins by mining 95,520 proteomics datasets. About 25% of the 7,264 ncORFs tested yielded detectable peptides. They introduced a classification framework that distinguishes true microproteins from 'peptideins' — ncORF products whose functional status remains uncertain. A new evolutionary metric, ORF relative branch length, revealed that many of these peptides show signs of evolutionary constraint, suggesting biological relevance. Follow-up experiments on one peptidein derived from the OLMALINC long non-coding RNA found it supports a core cellular function. Public databases including GENCODE and PeptideAtlas were updated with these results, offering researchers new tools to study a long-overlooked portion of the human proteome and expanding the catalog of potentially functional small proteins.

Read the original article →