Nutrient-Sensitive Reinforcement Learning in Monkeys.

In reinforcement learning (RL), animals choose by assigning values to options and learn by updating these values from reward outcomes. This framework has been instrumental in identifying fundamental learning variables and their neuronal implementations. However, canonical RL models do not explain how reward values are constructed from biologically critical intrinsic reward components, such as nutrients. From an ecological perspective, animals should adapt their foraging choices in dynamic environments to acquire nutrients that are essential for survival. Here, to advance the biological and ecological validity of RL models, we investigated how (male) monkeys adapt their choices to obtain preferred nutrient rewards under varying reward probabilities. We found that the nutrient composition of rewards strongly influenced learning and choices. Preferences of the animals for specific nutrients (sugar, fat) affected how they adapted to changing reward probabilities; the history of recent rewards influenced choices of the monkeys more strongly if these rewards contained the their preferred nutrients (nutrient-specific reward history). The monkeys also chose preferred nutrients even when they were associated with lower reward probability. A nutrient-sensitive RL model captured these processes; it updated the values of individual sugar and fat components of expected rewards based on experience and integrated them into subjective values that explained the choices of the monkeys. Nutrient-specific reward prediction errors guided this value-updating process. Our results identify nutrients as important reward components that guide learning and choice by influencing the subjective value of choice options. Extending RL models with nutrient-value functions may enhance their biological validity and uncover nutrient-specific learning and decision variables.SIGNIFICANCE STATEMENT RL is an influential framework that formalizes how animals learn from experienced rewards. Although reward is a foundational concept in RL theory, canonical RL models cannot explain how learning depends on specific reward properties, such as nutrients. Intuitively, learning should be sensitive to the nutrient components of the reward to benefit health and survival. Here, we show that the nutrient (fat, sugar) composition of rewards affects how the monkeys choose and learn in an RL paradigm and that key learning variables including reward history and reward prediction error should be modified with nutrient-specific components to account for the choice behavior observed in the monkeys. By incorporating biologically critical nutrient rewards into the RL framework, our findings help advance the ecological validity of RL models.

Assembly-inspired multiferroicity with nontrivial Chern insulating phase from exohedral metallofullerenes.

Fullerene-assembled low-dimensional materials have been experimentally realized in polymorphic forms and have attracted significant interest very recently. Here, we predict a two-dimensional (2D) honeycomb lattice material TM2(C60)3 (TM = Cr, Mo, and W) assembled from exohedral metallofullerene clusters TM(C60)3 that could exhibit planar triangular geometries. According to first-principles calculations combined with Monte Carlo simulations, we suggest that these 2D assembled materials exhibit various exotic physical properties, including ferromagnetism, ferroelectricity, and quantum anomalous Hall effect. Interestingly, mechanical strains could effectively tune their magnetic moments and switch the conducting spin channel of the Dirac bands at the Fermi level. Our work provides a new cluster-assembly design strategy toward cluster-assembled 2D materials based on fullerene characters.

Preferences for nutrients and sensory food qualities identify biological sources of economic values in monkeys.

Value is a foundational concept in reinforcement learning and economic choice theory. In these frameworks, individuals choose by assigning values to objects and learn by updating values with experience. These theories have been instrumental for revealing influences of probability, risk, and delay on choices. However, they do not explain how values are shaped by intrinsic properties of the choice objects themselves. Here, we investigated how economic value derives from the biologically critical components of foods: their nutrients and sensory qualities. When monkeys chose nutrient-defined liquids, they consistently preferred fat and sugar to low-nutrient alternatives. Rather than maximizing energy indiscriminately, they seemed to assign subjective values to specific nutrients, flexibly trading them against offered reward amounts. Nutrient-value functions accurately modeled these preferences, predicted choices across contexts, and accounted for individual differences. The monkeys' preferences shifted their daily nutrient balance away from dietary reference points, contrary to ecological foraging models but resembling human suboptimal eating in free-choice situations. To identify the sensory basis of nutrient values, we developed engineering tools that measured food textures on biological surfaces, mimicking oral conditions. Subjective valuations of two key texture parameters-viscosity and sliding friction-explained the monkeys' fat preferences, suggesting a texture-sensing mechanism for nutrient values. Extended reinforcement learning and choice models identified candidate neuronal mechanisms for nutrient-sensitive decision-making. These findings indicate that nutrients and food textures constitute critical reward components that shape economic values. Our nutrient-choice paradigm represents a promising tool for studying food-reward mechanisms in primates to better understand human-like eating behavior and obesity.

Artificial intelligence-enhanced white-light colonoscopy with attention guidance predicts colorectal cancer invasion depth.

BACKGROUND AND AIMS: Endoscopic submucosal dissection (ESD) and EMR are applied in treating superficial colorectal neoplasms but are contraindicated by deeply invasive colorectal cancer (CRC). The invasion depth of neoplasms can be examined by an automated artificial intelligence (AI) system to determine the applicability of ESD and EMR. METHODS: A deep convolutional neural network with a tumor localization branch to guide invasion depth classification was constructed on the GoogLeNet architecture. The model was trained using 7734 nonmagnified white-light colonoscopy (WLC) images supplemented by image augmentation from 657 lesions labeled with histopathologic analysis of invasion depth. An independent testing dataset consisting of 1634 WLC images from 156 lesions was used to validate the model. RESULTS: For predicting noninvasive and superficially invasive neoplasms, the model achieved an overall accuracy of 91.1% (95% confidence interval [CI], 89.6%-92.4%), with 91.2% sensitivity (95% CI, 88.8%-93.3%) and 91.0% specificity (95% CI, 89.0%-92.7%) at an optimal cutoff of .41 and the area under the receiver operating characteristic (AUROC) curve of .970 (95% CI, .962-.978). Inclusion of the advanced CRC data significantly increased the sensitivity in differentiating superficial neoplasms from deeply invasive early CRC to 65.3% (95% CI, 61.9%-68.8%) with an AUROC curve of .729 (95% CI, .699-.759), similar to experienced endoscopists (.691; 95% CI, .624-.758). CONCLUSIONS: We have developed an AI-enhanced attention-guided WLC system that differentiates noninvasive or superficially submucosal invasive neoplasms from deeply invasive CRC with high accuracy, sensitivity, and specificity.

Insulin secretion impairment induced by rosuvastatin partly though autophagy in INS-1E cells.

Statins are used extensively for the clinical treatment of cardiovascular diseases. Recent studies suggest that statins increase the risk of new-onset diabetes mellitus (NODM). However, the mechanisms of statin-induced NODM remain unclear. The present study investigated the effects of autophagy on insulin secretion impairment induced by rosuvastatin (RS) in rat insulinoma cells (INS-1E) cells. INS-1E cells were cultured and treated with RS at different concentrations (0.2-20 µM) for 24 h. Insulin secretion in INS-1E cells was detected by enzyme-linked immunosorbent assay, and the co-localization of microtubule-associated protein light chain 3 (LC3) and lysosome-associated membrane protein 2 (LAMP-2) was observed by immunofluorescence staining. Western blotting was used to assess the conversion of LC3 and p62. The results showed that the insulin secretion and cell viability decrease induced by RS treatment for 24 h occurred in a dose-dependent manner in INS-1E cells. RS significantly inhibited the expression of LC3-II but increased the protein expression of p62. Simultaneously, RS diminished the co-localization of LC3-II and LAMP-2 fluorescence signals. These results suggested that RS-inhibited autophagy in INS-1E cells. Rapamycin, an autophagy agonist, reversed the insulin secretion and cell viability suppression induced by RS in INS-1E cells. RS also decreased the phosphorylation of the mammalian target of rapamycin (mTOR). The results indicated that RS impairs insulin secretion in INS-1E cells, which may be partly due to the inhibition of autophagy via an mTOR-dependent pathway.

Stable self-compliance resistive switching in AlOδ/Ta2O(5-x)/TaOy triple layer devices.

Stable self-compliance property was observed in the AlOδ/Ta2O(5-x)/TaOy triple-layer resistive random access memory structure. The impact of AlOδ barrier layer was studied with different thicknesses. Endurance of more than 10(10) cycles and data retention for more than 3 h at 125 °C were demonstrated. All the measurements were carried out without external current compliance and no hard breakdown was observed. Systematic analysis reveals the self-compliance property is due to the built-in series resistance of the thin AlOδ barrier layer. A model is proposed to explain this self-compliance property.

Inhibition of kinetic random-distribution in DNA Seesaw gates and biosensors for complete leakage prevention.

DNA nanomaterials have a wide application prospect in biomedical field, among which DNA computers and biosensors based on Seesaw-based DNA circuit is considered to have the most development potential. However, the serious leakage of Seesaw-based DNA circuit prevented its further development and application. Moreover, the existing methods to suppress leakage can't achieve the ideal effect. Interestingly, we found a new source of leakage in Seesaw-based DNA circuit, which we think is the main reason why the previous methods to suppress leakage are not satisfactory. Therefore, based on this discovery, we use DNA triplex to design a new method to suppress the leakage of Seesaw-based DNA circuit. Its ingenious design makes it possible to perfectly suppress the leakage of all sources in Seesaw-based DNA circuit and ensure the normal output of the circuit. Based on this technology, we have constructed basic Seesaw module, AND gate, OR gate, secondary complex circuits and DNA detector. Experimental results show that we can increase the working range of the secondary Seesaw-based DNA circuit by five folds and keep its normal output signal above 90%, and we can improve the LOD of the Seesaw-based DNA detector to 1/11 of the traditional one(1.8pM). More importantly, we successfully developed a detector with adjustable detection range, which can theoretically achieve accurate detection in any concentration range. We believe the established triplex blocking strategy will greatly facilitate the most powerful Seesaw based DNA computers and biosensors, and further promote its application in biological systems.

Reference-free cell type deconvolution of multi-cellular pixel-resolution spatially resolved transcriptomics data.

Recent technological advancements have enabled spatially resolved transcriptomic profiling but at multi-cellular pixel resolution, thereby hindering the identification of cell-type-specific spatial patterns and gene expression variation. To address this challenge, we develop STdeconvolve as a reference-free approach to deconvolve underlying cell types comprising such multi-cellular pixel resolution spatial transcriptomics (ST) datasets. Using simulated as well as real ST datasets from diverse spatial transcriptomics technologies comprising a variety of spatial resolutions such as Spatial Transcriptomics, 10X Visium, DBiT-seq, and Slide-seq, we show that STdeconvolve can effectively recover cell-type transcriptional profiles and their proportional representation within pixels without reliance on external single-cell transcriptomics references. STdeconvolve provides comparable performance to existing reference-based methods when suitable single-cell references are available, as well as potentially superior performance when suitable single-cell references are not available. STdeconvolve is available as an open-source R software package with the source code available at https://github.com/JEFworks-Lab/STdeconvolve .

Achieving Significant Thermal Conductivity Enhancement via an Ice-Templated and Sintered BN-SiC Skeleton.

Conventional polymer composites normally suffer from undesired thermal conductivity enhancement which has hampered the development of modern electronics as they face a stricter heat dissipating requirement. It is still challenging to achieve satisfactory thermal conductivity enhancement with reasonable mechanical properties. Herein, we present a three-dimensional (3D), lightweight, and mechanically strong boron nitride (BN)-silicon carbide (SiC) skeleton with aligned thermal pathways via the combination of ice-templated assembly and high-temperature sintering. The sintering has introduced atomic-level coupling at the BN-SiC junction which contributes to efficient phonon transport via the newly formed borosilicate glass BCxN3-x (0 ≤ x ≤ 3) and SiCxN4-x (0 ≤ x ≤ 4) phases, leading to much lower interfacial thermal resistance. Thus, the obtained BN-SiC skeleton shows satisfactory thermal performance. The prepared 3D BN-SiC/polydimethylsiloxane (PDMS) composites exhibit a maximum through-plane thermal conductivity of 3.87 W·m-1·K-1 at a filler loading of only 8.35 vol %. The thermal conductivity enhancement efficiency reaches 220% per 1 vol % filler when compared to pure PDMS matrix, superior to other reported BN skeleton-based composites. The feature of our strategy is to allow the oriented three-dimensional skeleton to be strongly bonded by a sintered ceramic phase instead of polymer-like adhesive, namely, to improve the intrinsic thermal conductivity of the skeleton to the greatest extent. This strategy can be applied to develop novel thermal management materials that are lightweight and mechanically tough that rapidly transfer heat. It represents a new avenue to addressing the heat challenges in traditional electronic products.

Roles and Regulation of Long Noncoding RNAs in Hepatocellular Carcinoma.

Next-generation sequencing has uncovered thousands of long noncoding RNAs (lncRNA). Many are reported to be aberrantly expressed in various cancers, including hepatocellular carcinoma (HCC), and play key roles in tumorigenesis. This review provides an in-depth discussion of the oncogenic mechanisms reported to be associated with deregulated HCC-associated lncRNAs. Transcriptional expression of lncRNAs in HCC is modulated through transcription factors, or epigenetically by aberrant histone acetylation or DNA methylation, and posttranscriptionally by lncRNA transcript stability modulated by miRNAs and RNA-binding proteins. Seventy-four deregulated lncRNAs have been identified in HCC, of which, 52 are upregulated. This review maps the oncogenic roles of these deregulated lncRNAs by integrating diverse datasets including clinicopathologic features, affected cancer phenotypes, associated miRNA and/or protein-interacting partners as well as modulated gene/protein expression. Notably, 63 deregulated lncRNAs are significantly associated with clinicopathologic features of HCC. Twenty-three deregulated lncRNAs associated with both tumor and metastatic clinical features were also tumorigenic and prometastatic in experimental models of HCC, and eight of these mapped to known cancer pathways. Fifty-two upregulated lncRNAs exhibit oncogenic properties and are associated with prominent hallmarks of cancer, whereas 22 downregulated lncRNAs have tumor-suppressive properties. Aberrantly expressed lncRNAs in HCC exert pleiotropic effects on miRNAs, mRNAs, and proteins. They affect multiple cancer phenotypes by altering miRNA and mRNA expression and stability, as well as through effects on protein expression, degradation, structure, or interactions with transcriptional regulators. Hence, these insights reveal novel lncRNAs as potential biomarkers and may enable the design of precision therapy for HCC.