Tuesday, October 27
234 new papers appeared. We read every one. These 16 are worth your time.
Today's edition
Picked for broad interest, a real step forward and honest claims, with no field allowed to take over.
Design of full-length agonist and antagonist antibodies for hard targets
To our knowledge, this study provides the first demonstration of both GPCR agonism and antagonism by de novo designed full-length IgG antibodies.
AI summary
The authors built a computer pipeline called Pythia-1 that designs full-length antibodies from scratch. They tested 22 hard drug targets including GPCRs, a class that is especially difficult for antibody discovery. Screening at most 48 designs per target found antibodies with single-digit nanomolar binding strength. The pipeline designed antibodies that activate or block target proteins, including the first de novo full-length IgG antibodies that both activate and block GPCRs. The designs also met practical requirements for making drugs, with quality matching or exceeding clinical-stage antibodies.
Written by Kimi K2.5 from the abstract only. Jev checked every sentence against the authors' words. See where each sentence comes from
PreprintClaims a big stepReal-world useCulture-free Helicobacter pylori genomics reveals within-host virulence heterogeneity and antibiotic resistance
Here we develop ultra-sensitive Hp sequencing (uHp-seq) for reconstructing Hp genomes directly from gastric biopsies. uHp-seq achieves a thousand-fold Hp DNA enrichment and recovers near-complete genomes across samples spanning orders of magnitude in Hp abundance.
AI summary
The researchers developed a method called uHp-seq that can read the complete genetic code of Helicobacter pylori bacteria directly from stomach tissue samples. This method enriches Hp DNA a thousand times and works even when the bacteria are extremely scarce in the sample. They found that a single person can carry multiple Hp strains with different cancer-causing potential at the same time. The method detects antibiotic resistance variants more accurately than existing commercial tests.
Written by Kimi K2.5 from the abstract only. Jev checked every sentence against the authors' words. See where each sentence comes from
PreprintReal-world useSequential vaccination matures VRC01-class HIV-1 neutralizing antibodies in humans
This work advances the HIV-1 germline-targeting vaccination concept from proof-of-priming towards proof-of-breadth and defines a benchmark for bnAb induction.
AI summary
In a human trial, they gave a second vaccine shot to people who had received an earlier HIV vaccine. This reactivated memory immune cells and improved the antibodies' ability to bind to a key site on HIV. Three antibodies from different cell lineages neutralized 18-31% of 119 diverse HIV viruses. The study moves HIV vaccine research from showing priming works toward showing it can produce broad protection.
Written by Kimi K2.5 from the abstract only. Jev checked every sentence against the authors' words. See where each sentence comes from
PreprintReal-world useExhaustive white blood cell extraction and profiling across whole-slide images of blood and bone marrow smears
Here, we present HemAutomaton, a framework built on neural cellular automata that extracts white blood cells across entire whole-slide images without manual region selection and with minimal annotation requirements.
AI summary
Blood and bone marrow smears are used to diagnose blood disorders. The authors created HemAutomaton, a neural network system that finds white blood cells across entire scanned slides without manual selection. HemAutomaton matches Faster R-CNN accuracy but uses far less computing power and over 10,000 times fewer parameters. The system extracted over 100 million single-cell images from three blood cancer groups and performed efficient cell counting using few-shot classification. Automated blast counts matched expert evaluations within normal expert disagreement ranges.
Written by Kimi K2.5 from the abstract only. Jev checked every sentence against the authors' words. See where each sentence comes from
PreprintReal-world useCornealia: automated collective order analysis reveals density-dependent reorganization of donor corneal endothelium
Across 2,177 donor cornea images from two cohorts, collective order, distinct from classical metrics, responds most to mechanical perturbation: corneas with endothelial cell density above 2,400 cells/mm^2 mostly gain order, whereas less dense ones lose it, on one curve shared by both cohorts.
AI summary
The authors created Cornealia, an automated tool to analyze how corneal cells are organized together and the forces between them. In lab-grown cells, this collective organization dropped when cells were repeatedly grown and was higher when growth factors kept their connections intact. In 2,177 donor cornea images, denser corneas above 2,400 cells per square millimeter mostly gained organization when disturbed, while less dense ones lost it. This switch occurs within the minimum donor densities eye banks use, suggesting these clinical thresholds have a mechanical basis.
Written by Kimi K2.5 from the abstract only. Jev checked every sentence against the authors' words. See where each sentence comes from
PreprintReal-world useSimulating Psychotherapeutic Sessions with Generative Artificial Intelligence: A Proof-of-Concept Study of In Silico Psychotherapy Research
Conclusions: LLM-based psychotherapy simulations were technically feasible, demonstrated experimental fidelity, and showed theoretically concordant effects beyond directly manipulated constructs and simulation participants.
AI summary
The authors used AI to simulate 512 therapy sessions between virtual therapists and patients. Sessions with high empathy scored better on treatment integrity scales (beta = 0.54-0.95, p<0.001). Simulated patients also showed increased confidence for change (beta = 0.17, p<0.001). The authors conclude this AI simulation method may help test therapy ideas before real patient trials.
Written by Kimi K2.5 from the abstract only. Jev checked every sentence against the authors' words. See where each sentence comes from
PreprintRepetitive transcranial magnetic stimulation targeting motor cortical region induces cell-type-specific transcriptional reprogramming and behavioral recovery in non-human primate model of Rett syndrome
We identified a new treatment target, ventral premotor cortex (PMv), for RTT and found that following a 20-day rTMS course, several behavioral and functional connectivity measures showed directional changes.
AI summary
Rett syndrome is a severe brain development disorder caused by mutations in the MECP2 gene. The researchers tested repetitive magnetic brain stimulation on monkeys with this syndrome. The treatment also fixed abnormal gene activity in specific brain cell types. The authors say this brain stimulation approach may treat Rett syndrome, but needs more testing.
Written by Kimi K2.5 from the abstract only. Jev checked every sentence against the authors' words. See where each sentence comes from
PreprintReal-world useEye movements enhance stereoscopic 3D vision
Here we show that stereopsis critically depends on these oculomotor-induced input changes.
AI summary
Stereopsis is the brain's ability to judge depth from the slight differences between what each eye sees. Tiny eye movements constantly change these differences, making them vary over time. The authors tracked both eyes with high precision and controlled what each eye saw based on gaze position. The amount of depth change from eye movements predicted how well people saw depth, no matter which type of eye movement caused it.
Written by Kimi K2.5 from the abstract only. Jev checked every sentence against the authors' words. See where each sentence comes from
PreprintLesions causing room tilt illusion link a cerebello-cingulate brain network to verticality perception
In conclusion, anatomically diverse RTI lesions localize to a specific cerebello-cingulate network, providing an anatomical framework for verticality perception in the human brain.
AI summary
The authors studied 37 patients whose brain lesions caused room tilt illusion, a condition where the visual world suddenly seems to flip. They found these lesions were scattered across many brain areas, but most often in the cerebellum. Despite different locations, the lesions all connected to two hub regions: the posterior cerebellum and the mid-cingulate cortex. The results suggest a specific cerebello-cingulate network supports our sense of vertical orientation.
Written by Kimi K2.5 from the abstract only. Jev checked every sentence against the authors' words. See where each sentence comes from
PreprintArabian hydroclimate and hominin population dynamics over the last 500,000 years
This study reveals potential hominin dispersals across the Arabian Peninsula in both traditional humid periods and brief glacial wet spells from a marine dust dynamics perspective.
AI summary
Researchers analyzed dust from the western Indian Ocean covering the past 500,000 years. Human populations in Arabia grew during long wet periods. Short bursts of heavy rain during ice ages also allowed humans to expand there. Humans may have spread across Arabia during both traditional wet periods and brief glacial wet spells.
Written by Kimi K2.5 from the abstract only. Jev checked every sentence against the authors' words. See where each sentence comes from
Peer-reviewed journalStop guessing thinning distances: a data-driven framework for spatial thinning parametrization for ecological modelling
Here, we propose a data-driven framework for spatial thinning parameterization that replaces heuristic decision-making with principled, reproducible criteria.
AI summary
Researchers usually pick spatial thinning parameters using rough rules of thumb, which adds unmeasured error. One method finds the largest thinning that keeps species niche estimates stable; the other scales thinning by range size. Testing with over 100,000 seagrass records showed optimal thinning varied with range size, and uniform approaches lost too much data or left clustering. The framework also flagged species whose data could not support stable niche estimates, even with minimal thinning.
Written by Kimi K2.5 from the abstract only. Jev checked every sentence against the authors' words. See where each sentence comes from
PreprintOne-shot synthesis of rare gastrointestinal lesions improves diagnostic accuracy and clinical training
Here we present EndoRare, a one-shot, retraining-free generative framework that synthesizes diverse lesion exemplars from a single reference image.
AI summary
Rare gastrointestinal lesions are hard to study because doctors rarely see them during routine exams. The authors created EndoRare, a system that makes diverse images of rare lesions from just one example photo. EndoRare separates key disease features from irrelevant details to create varied but realistic images. Experts found the synthetic images believable, and using them to train AI improved detection accuracy. Novice doctors who saw EndoRare cases also showed better recognition of these rare lesions.
Written by Kimi K2.5 from the abstract only. Jev checked every sentence against the authors' words. See where each sentence comes from
Peer-reviewed journalReal-world useMetagenomic evidence for climate-associated variation in microbial methylmercury production across peatlands
Together, these findings indicate that warmer climates may be associated with microbial community restructuring and lower net MeHg production.
AI summary
Peatlands store large amounts of mercury and are sensitive to warming. Researchers analyzed 144 peat soil samples to study how climate affects microbial production of methylmercury. They found more mercury-methylation genes in colder regions and more demethylation genes in warmer regions. Nitrogen cycling genes were linked to mercury transformation processes. The findings suggest that warmer climates may lead to lower net methylmercury production.
Written by Kimi K2.5 from the abstract only. Jev checked every sentence against the authors' words. See where each sentence comes from
Peer-reviewed journalReal-world useEvaluating anonymized genome re-identification using polygenic predictions and its implications for data privacy
Thus, although re-identification by phenotypic prediction is technically feasible, its effectiveness in real-world conditions remains limited.
AI summary
The authors developed a method to test if someone's genome can be matched to their known traits using genetic predictions. Their method works well in theory, reaching over 99% precision when assuming a 50% chance the genome and traits belong to the same person. The attack also fails to reliably reveal sensitive genetic information like Alzheimer's risk. The authors conclude that while this re-identification method is technically possible, it remains limited in real-world conditions.
Written by Kimi K2.5 from the abstract only. Jev checked every sentence against the authors' words. See where each sentence comes from
Peer-reviewed journalEmergence of diffusion models from associative memory through memorization and generalization
We identify a critical intermediate phase, predicted by DenseAM theory, characterized by the emergence of spurious states.
AI summary
Dense Associative Memories store training data as energy landscape valleys. When training data exceeds capacity, new valleys appear that differ from training data. These extra valleys, called spurious states, can block memory retrieval. The authors study diffusion models as memory retrieval attempts. With small data, diffusion models make separate attractors for each training sample. As data grows, diffusion models shift to generalization. A middle phase emerges where spurious states appear, as theory predicts. In generative modeling, these states are not flaws but first signs of generation ability. The authors show these states exist across many architectures and datasets.
Written by Kimi K2.5 from the abstract only. Jev checked every sentence against the authors' words. See where each sentence comes from
Peer-reviewed journalIdentification of insecticidal double-stranded RNAs targeting the proteasome pathway in the hop flea beetle Psylliodes attenuata using dsRIP
Overall, the feeding bioassays and molecular resources establish P. attenuata as a promising target for RNAi-based pest control and demonstrate the usefulness of dsRIP for RNAi research in insects previously unstudied at the molecular level.
AI summary
The hop flea beetle is a major pest of hops, the plant used to make beer. Researchers used a web platform called dsRIP to find genes they could target with double-stranded RNA. Three designed RNAs targeting proteasome parts caused up to 64% death and 45% feeding reduction in feeding tests. Molecular tests confirmed the RNA interference pathway was activated in the beetles. The authors conclude this beetle is a promising target for RNA-based pest control.
Written by Kimi K2.5 from the abstract only. Jev checked every sentence against the authors' words. See where each sentence comes from
PreprintReal-world use