bioRxiv : the preprint server for biology最新文献

The "holy grail" of chromatin research would be to follow the chromatin configuration in individual live cells over time. One way to achieve this goal would be to track the positions of multiple loci arranged along the chromatin polymer with fluorescent labels. Use of distinguishable labels would define each locus uniquely in a microscopic image but would restrict the number of loci that could be observed simultaneously, because of experimental limits to the number of distinguishable labels. Use of the same label for all loci circumvents this limitation but requires a (currently lacking) framework for how to establish each observed locus identity, i.e. to which genomic position it corresponds. Here we analyze theoretically, using simulations of Rouse-model polymers, how single-particle-tracking of multiple identically-labeled loci enables determination of loci identity. We show that the probability of correctly assigning observed loci to genomic positions converges exponentially to unity as the number of observed loci configurations increases. The convergence rate depends only weakly on the number of labeled loci, so that even large numbers of loci can be identified with high fidelity by tracking them across about 8 independent chromatin configurations. In the case of two distinct labels that alternate along the chromatin polymer, we find that the probability of the correct assignment converges faster than for same-labeled loci, requiring observation of fewer independent chromatin configurations to establish loci identities. Finally, for a modified Rouse-model polymer, that realizes a population of dynamic loops, we find that the success probability also converges to unity exponentially as the number of observed loci configurations increases, albeit slightly more slowly than for a classical Rouse model polymer. Altogether, these results establish particle tracking of multiple identically- or alternately-labeled loci over time as a feasible way to infer temporal dynamics of the coarse-grained configuration of the chromatin polymer in individual living cells.

There is growing interest in pre-training antibody language models (AbLMs) with a mixture of unpaired and natively paired sequences, seeking to combine the proven benefits of training with natively paired sequences with the massive scale of unpaired antibody sequence datasets. However, given the novelty of this strategy, the field lacks a systematic evaluation of data processing methods and training strategies that maximize the benefits of mixed training data while accommodating the significant imbalance in the size of existing paired and unpaired datasets. Here we introduce a method of curriculum learning for AbLMs, which facilitates a gradual transition from unpaired to paired sequences during training. We optimize this method and show that a 650M-parameter curriculum model, CurrAb, outperforms existing mixed AbLMs in downstream classification tasks.

Representations of chemical features in the neural activity of the olfactory bulb (OB) are not well-understood, unlike the neural code for stimuli of the other sensory modalities. This is because the space of olfactory stimuli lacks a natural coordinate system, and this significantly complicates characterizing neural receptive fields (tuning curves), analogous to those in the other sensory modalities. The degree to which olfactory tuning is spatially organized across the OB, often referred to as chemotopy, is also not well-understood. To advance our understanding of these aspects of olfactory coding, we introduce an interpretable method of Shapley fields, as an olfactory analog of retinotopic receptive fields. Shapley fields are spatial distributions of chemical feature importance for the tuning of OB glomeruli. We used this tool to investigate chemotopy in the OB with diverse sets of chemical features using widefield epifluorescence recordings of the mouse dorsal OB in response to stimuli across a wide range of the chemical space. We found that Shapley fields reveal a weak chemotopic organization of the chemical feature sensitivity of dorsal OB glomeruli. This organization was consistent across animals and mostly agreed across very different chemical feature sets: (i) the expert-curated PubChem database features and (ii) features derived from a Graph Neural Network trained on human olfactory perceptual tasks. Moreover, we found that the principal components of the Shapley fields often corresponded to single commonly accepted chemical classification groups, that therefore could be "recovered" from the neural activity in the mouse OB. Our findings suggest that Shapley fields may serve as a chemical feature-agnostic method for investigating olfactory perception.

Understanding how cellular pathways interact is crucial for treating complex diseases like cancer, yet our ability to map these connections systematically remains limited. Individual gene-gene interaction studies have provided insights ^1,2 , but they miss the emergent properties of pathways working together. To address this challenge, we developed a multi-gene approach to pathway mapping and applied it to CRISPR data from the Cancer Dependency Map ³ . Our analysis of the electron transport chain revealed certain blood cancers, including acute myeloid leukemia (AML), depend on an unexpected link between Complex II and purine metabolism. Through stable isotope metabolomic tracing, we found that Complex II directly supports de novo purine biosynthesis and exogenous purines rescue AML from Complex II inhibition. The mechanism involves a metabolic circuit where glutamine provides nitrogen to build the purine ring, producing glutamate that Complex II must oxidize to sustain purine synthesis. This connection translated to a metabolic vulnerability whereby increasing intracellular glutamate levels suppresses purine production and sensitizes AML to Complex II inhibition. In mouse models, targeting Complex II triggered rapid disease regression and extended survival in aggressive AML. The clinical relevance of this pathway emerged in human studies, where higher Complex II gene expression correlates with both resistance to mitochondria-targeted therapies and worse survival outcomes specifically in AML patients. These findings establish Complex II as a central regulator of de novo purine biosynthesis and identify it as a promising therapeutic target in AML.

Neurons exhibit extraordinary precision in selecting synaptic partners. Whereas cell-surface proteins (CSPs) mediating attractive interactions between developing axons and dendrites have been shown to instruct synaptic partner matching ^1,2 , it is less clear the degree to which repulsive interactions play a role. Here, using a genetic screen guided by single cell transcriptomes ^3,4 , we identified three CSP pairs-Toll2-Ptp10D, Fili-Kek1, and Hbs/Sns-Kirre-in mediating repulsive interactions between non-partner olfactory receptor neuron (ORN) axons and projection neuron (PN) dendrites in the developing Drosophila olfactory circuit. Each CSP pair exhibits inverse expression patterns in the select PN-ORN partners. Loss of each CSP in ORNs led to similar synaptic partner matching deficits as the loss of its partner CSP in PNs, and mistargeting phenotypes caused by overexpressing one CSP could be suppressed by loss of its partner CSP. Each CSP pair is also differentially expressed in other brain regions. Together, our data reveal that multiple repulsive CSP pairs work together to ensure precise synaptic partner matching during development by preventing neurons from forming connections with non-cognate partners.

To correctly parse the visual scene, one must detect edges and determine their underlying cause. Previous work has demonstrated that image-computable neural networks trained to differentiate natural shadow and occlusion edges exhibited sensitivity to boundary sharpness and texture differences. Although these models showed a strong correlation with human performance on an edge classification task, this previous study did not directly investigate whether humans actually make use of boundary sharpness and texture cues when classifying edges as shadows or occlusions. Here we directly investigated this using synthetic image patch stimuli formed by quilting together two different natural textures, allowing us to parametrically manipulate boundary sharpness, texture modulation, and luminance modulation. In a series of initial "training" experiments, observers were trained to correctly identify the cause of natural image patches taken from one of three categories (occlusion, shadow, uniform texture). In a subsequent series of "test" experiments, these same observers then classified 5 sets of synthetic boundary images defined by varying boundary sharpness, luminance modulation, and texture modulation cues using a set of novel parametric stimuli. These three visual cues exhibited strong interactions to determine categorization probabilities. For sharp edges, increasing luminance modulation made it less likely the patch would be classified as a texture and more likely it would be classified as an occlusion, whereas for blurred edges, increasing luminance modulation made it more likely the patch would be classified as a shadow. Boundary sharpness had a profound effect, so that in the presence of luminance modulation increasing sharpness decreased the likelihood of classification as a shadow and increased the likelihood of classification as an occlusion. Texture modulation had little effect on categorization, except in the case of a sharp boundary with zero luminance modulation. Results were consistent across all 5 stimulus sets, showing these effects are not due to the idiosyncrasies of the particular texture pairs. Human performance was found to be well explained by a simple linear multinomial logistic regression model defined on luminance, texture and sharpness cues, with slightly improved performance for a more complicated nonlinear model taking multiplicative parameter combinations into account. Our results demonstrate that human observers make use of the same cues as our previous machine learning models when detecting edges and determining their cause, helping us to better understand the neural and perceptual mechanisms of scene parsing.

The dynamics of antimicrobial resistance in bacterial populations are informed by the fitness impact of genetic determinants of resistance and antibiotic pressure. However, estimates of real-world fitness impact have been lacking. To address this gap, we developed a hierarchical Bayesian phylodynamic model to quantify contributions of resistance determinants to strain success in a 20-year collection of Neisseria gonorrhoeae isolates. Fitness contributions varied with antibiotic use, and genetic pathways to phenotypically identical resistance conferred distinct fitness effects. These findings were supported by in vitro and experimental infection competition. Quantifying these fitness contributions to lineage dynamics reveals opportunities for investigation into other genetic and environmental drivers of fitness. This work thus establishes a method for linking pathogen genomics and antibiotic use to define factors shaping ecological trends.

The structural dynamics of proteins play a crucial role in their function, yet most experimental and deep learning methods produce only static models. While molecular dynamics (MD) simulations provide atomistic insight into conformational transitions, they remain computationally prohibitive, particularly for large-scale motions. Here, we introduce DeepPath, a deep-learning-based framework that rapidly generates physically realistic transition pathways between known protein states. Unlike conventional supervised learning approaches, DeepPath employs active learning to iteratively refine its predictions, leveraging molecular mechanical force fields as an oracle to guide pathway generation. We validated DeepPath on three biologically relevant test cases: SHP2 activation, CdiB H1 secretion, and the BAM complex lateral gate opening. DeepPath accurately predicted the transition pathways for all test cases, reproducing key intermediate structures and transient interactions observed in previous studies. Notably, DeepPath also predicted an intermediate between the BAM inward- and outward-open states that closely aligns with an experimentally observed hybrid-barrel structure (TMscore = 0.91). Across all cases, DeepPath achieved accurate pathway predictions within hours, showcasing an efficient alternative to MD simulations for exploring protein conformational transitions.

Bacteria and cancer cells live in a spatially heterogeneous environment, where migration shapes the microhabitat structures critical for colonization and metastasis. The interplay between growth, migration, and microhabitat structure complicates the prediction of population responses to drugs, such as clearance or sustained growth, posing a longstanding challenge. Here, we disentangle growth-migration dynamics and identify that population decline is determined by two decoupled terms: a spatial growth variation term and a microhabitat structure term. Notably, the microhabitat structure term can be interpreted as a dynamic-related centrality measure. For fixed spatial drug arrangements, we show that interpreting these centralities reveals how different network structures, even with identical edge densities, microhabitat numbers, and spatial heterogeneity, can lead to distinct population-level responses. Increasing edge density shifts the population response from growth to clearance, supporting an inversed centrality-connectivity relationship, and mirroring the effects of higher migration rates. Furthermore, we derive a sufficient condition for robust population decline across various spatial growth rate arrangements, regardless of spatial-temporal fluctuations induced by drugs. Additionally, we demonstrate that varying the maximum growth-to-death ratio, determined by drug-bacteria interactions, can lead to distinct population decline profiles and a minimal decline phase emerges. These findings address key challenges in predicting population-level responses and provide insights into divergent clinical outcomes under identical drug dosages. This work may offer a new method of interpreting treatment dynamics and potential approaches for optimizing spatially explicit drug dosing strategies.

Drugs of abuse in adolescence impact brain maturation and increase psychiatric risk, with differences in sensitivity between males and females. Amphetamine in adolescent male, but not female mice, causes dopamine axons intended to innervate the nucleus accumbens and to grow ectopically to the prefrontal cortex (PFC). This is mediated by drug-induced downregulation of the Netrin-1 receptor DCC. How off-target dopamine axons function in the adult PFC remains to be determined. Here we report that males and females show place preference for amphetamine in adolescence. However, only in males, amphetamine increases PFC dopamine transporter expression in adulthood: leading to aberrant baseline dopamine transients, faster dopamine release, and exaggerated responses to acute methylphenidate. Upregulation of DCC in adolescence, using CRISPRa, prevents all these changes. Mesolimbic dopamine axons rerouted to the PFC in adolescence retain anatomical and functional phenotypes of their intended target, rendering males enduringly vulnerable to the harmful effects of drugs of abuse.