Fundamental Research最新文献

Childhood asthma is one of the most common respiratory diseases with rising mortality and morbidity. The multi-omics data is providing a new chance to explore collaborative biomarkers and corresponding diagnostic models of childhood asthma. To capture the nonlinear association of multi-omics data and improve interpretability of diagnostic model, we proposed a novel deep association model (DAM) and corresponding efficient analysis framework. First, the Deep Subspace Reconstruction was used to fuse the omics data and diagnostic information, thereby correcting the distribution of the original omics data and reducing the influence of unnecessary data noises. Second, the Joint Deep Semi-Negative Matrix Factorization was applied to identify different latent sample patterns and extract biomarkers from different omics data levels. Third, our newly proposed Deep Orthogonal Canonical Correlation Analysis can rank features in the collaborative module, which are able to construct the diagnostic model considering nonlinear correlation between different omics data levels. Using DAM, we deeply analyzed the transcriptome and methylation data of childhood asthma. The effectiveness of DAM is verified from the perspectives of algorithm performance and biological significance on the independent test dataset, by ablation experiment and comparison with many baseline methods from clinical and biological studies. The DAM-induced diagnostic model can achieve a prediction AUC of 0.912, which is higher than that of many other alternative methods. Meanwhile, relevant pathways and biomarkers of childhood asthma are also recognized to be collectively altered on the gene expression and methylation levels. As an interpretable machine learning approach, DAM simultaneously considers the non-linear associations among samples and those among biological features, which should help explore interpretative biomarker candidates and efficient diagnostic models from multi-omics data analysis for human complex diseases.

Neural network pruning is a popular approach to reducing the computational complexity of deep neural networks. In recent years, as growing evidence shows that conventional network pruning methods employ inappropriate proxy metrics, and as new types of hardware become increasingly available, hardware-aware network pruning that incorporates hardware characteristics in the loop of network pruning has gained growing attention. Both network accuracy and hardware efficiency (latency, memory consumption, etc.) are critical objectives to the success of network pruning, but the conflict between the multiple objectives makes it impossible to find a single optimal solution. Previous studies mostly convert the hardware-aware network pruning to optimization problems with a single objective. In this paper, we propose to solve the hardware-aware network pruning problem with Multi-Objective Evolutionary Algorithms (MOEAs). Specifically, we formulate the problem as a multi-objective optimization problem, and propose a novel memetic MOEA, namely HAMP, that combines an efficient portfolio-based selection and a surrogate-assisted local search, to solve it. Empirical studies demonstrate the potential of MOEAs in providing simultaneously a set of alternative solutions and the superiority of HAMP compared to the state-of-the-art hardware-aware network pruning method.

Drug discovery is costly and time consuming, and modern drug discovery endeavors are progressively reliant on computational methodologies, aiming to mitigate temporal and financial expenditures associated with the process. In particular, the time required for vaccine and drug discovery is prolonged during emergency situations such as the coronavirus 2019 pandemic. Recently, the performance of deep learning methods in drug virtual screening has been particularly prominent. It has become a concern for researchers how to summarize the existing deep learning in drug virtual screening, select different models for different drug screening problems, exploit the advantages of deep learning models, and further improve the capability of deep learning in drug virtual screening. This review first introduces the basic concepts of drug virtual screening, common datasets, and data representation methods. Then, large numbers of common deep learning methods for drug virtual screening are compared and analyzed. In addition, a dataset of different sizes is constructed independently to evaluate the performance of each deep learning model for the difficult problem of large-scale ligand virtual screening. Finally, the existing challenges and future directions in the field of virtual screening are presented.

Since their clinical application in the 1840s, the greatest mystery surrounding general anesthesia (GA) is how different kinds of general anesthetics cause reversible unconsciousness, and the precise neural mechanisms underlying the processes. Over past years, although many studies revealed the roles of cortex, thalamus, brainstem, especially the sleep-wake circuits in GA-induced loss of consciousness (LOC),the full picture of the neural circuit mechanism of GA is still largely unknown. Recent studies have focused on the importance of other brain regions. Here, we report that the activity of glutamatergic (Glu) neurons in the piriform cortex (PC), a critical brain region for odor encoding, began to increase during the LOC of GA and gradually recovered after recovery of consciousness. Chemical lesions of the anterior PC (APC) neurons accelerated the induction time of isoflurane anesthesia. Chemogenetic and optogenetic activation of APC^Glu neurons prolonged isoflurane and sevoflurane anesthesia induction, whereas APC^Glu neuron inhibition displayed the opposite effects. Moreover, the modification of APC^Glu neurons did not affect the induction or emergence time of propofol GA. In addition, odor processing may be partially involved in the induction of isoflurane and sevoflurane GA regulated by APC^Glu neurons. In conclusion, our findings reveal a critical role of APC^Glu neurons in inhalational GA induction.

Quantum secure direct communication (QSDC) can directly transmit secret messages through quantum channel without keys. Device-independent (DI) QSDC guarantees the message security relying only on the observation of the Bell-inequality violation, but not on any detailed description or trust of the devices’ inner workings. Compared with conventional QSDC, DI-QSDC has relatively low secret message capacity. To increase DI-QSDC’s secret messages capacity, we propose a high-capacity DI-QSDC protocol based on the hyper-encoding technique. The total message leakage rate of our DI-QSDC protocol only relies on the most robust degree of freedom. We provide the numerical simulation of its secret message capacity altered with the communication distance. Our work serves as an important step toward the further development of DI-QSDC systems.

The potential for being able to identify individuals at high disease risk solely based on genotype data has garnered significant interest. Although widely applied, traditional polygenic risk scoring methods fall short, as they are built on additive models that fail to capture the intricate associations among single nucleotide polymorphisms (SNPs). This presents a limitation, as genetic diseases often arise from complex interactions between multiple SNPs. To address this challenge, we developed DeepRisk, a biological knowledge-driven deep learning method for modeling these complex, nonlinear associations among SNPs, to provide a more effective method for scoring the risk of common diseases with genome-wide genotype data. Evaluations demonstrated that DeepRisk outperforms existing PRS-based methods in identifying individuals at high risk for four common diseases: Alzheimer's disease, inflammatory bowel disease, type 2 diabetes, and breast cancer.

The genome-wide association study (GWAS) aims to detect associations between individual single nucleotide polymorphisms (SNPs) or SNP interactions and phenotypes to decipher the genetic mechanism. Existing GWAS analysis tools have different focuses and advantages, but suffer a series of tedious and heterogeneous configurations for computation. It is inconvenient for researchers to simply choose and apply these tools, statistically and biologically analyze their results for different usages. To address these issues, we develop a user friendly web pipeline GWASTool for detecting associations, which includes simulation data generation, associated loci detection, result visualization, analysis and comparison. GWASTool provides a unified and plugin-able framework to encapsulate the heterogeneity of GWAS algorithms, simplifies the analysis steps and energizes GWAS tasks. GWASTool is implemented in Java and is freely available for public use at http://www.sdu-idea.cn/GWASTool. The website hosts a comprehensive collection of resources, including a user manual, description of integrated algorithms, data examples and standalone version for download.

In addition to the cardinal motor symptoms, pain is a major non-motor symptom of Parkinson's disease (PD). Neuroinflammation in the substantia nigra pars compacta and dorsal striatum is involved in neurodegeneration in PD. But the polarization of microglia and astrocytes in the dorsal striatum and their contribution to motor deficits and hyperalgesia in PD have not been characterized. In the present study, we observed that hemiparkinsonian mice established by unilateral 6-OHDA injection in the medial forebrain bundle exhibited motor deficits and mechanical allodynia. In these mice, both microglia and astrocytes in the dorsal striatum were activated and polarized to M1/M2 microglia and A1/A2 astrocytes as genes specific to these cells were upregulated. These effects peaked 7 days after 6-OHDA injection. Meanwhile, striatal astrocytes in parkinsonian mice also displayed hyperpolarized membrane potentials, enhanced voltage-gated potassium currents, and dysfunction in inwardly rectifying potassium channels and glutamate transporters. Systemic administration of minocycline, a microglia inhibitor, attenuated the expression of genes specific to M1 microglia and A1 astrocytes in the dorsal striatum (but not those specific to M2 microglia and A2 astrocytes), attenuated the damage in the nigrostriatal dopaminergic system, and alleviated the motor deficits and mechanical allodynia in parkinsonian mice. By contrast, local administration of minocycline into the dorsal striatum of parkinsonian mice mitigated only hyperalgesia. This study suggests that M1 microglia and A1 astrocytes in the dorsal striatum may play important roles in the development of pathophysiology underlying hyperalgesia in the early stages of PD.