中国地方品种鸡的分类研究

中国鸡种资源极为丰富,中国历史上形成的许多优良地方品种鸡,对中国养鸡事业发展和世界养鸡业都曾做出过重要贡献。为充分利用我国地方鸡种资源遗传潜力和杂种优势,本文利用图论主成分分类法与系统聚类法根据20项指标研究了中国30个地方品种鸡的分类问题,分类结果比较符合实际,特别是图论主成分分类图直观清楚,生物学含意明确,对中国鸡种资源和基因库的利用有实际参考价值。     

Coupling Molecular Dynamics and Deep Learning to Mine Protein Conformational Space

1. Download : Download high-res image (302KB)
2. Download : Download full-size image

     

Dynamic graph convolution neural network based on spatial-temporal correlation for air quality prediction

Air pollution is a serious threat to both the ecological environment and the physical health of individuals. Therefore, accurate air quality prediction is urgent and necessary for pollution mitigation and residents’ travel. However, few existing models are established based on the dynamic spatiotemporal correlation of air pollutants to predict air quality. In this paper, a novel deep learning model combining the dynamic graph convolutional network and the multi-channel temporal convolutional network (DGC-MTCN) is proposed for air quality prediction. To efficiently represent the time-varying spatial dependencies, a new spatiotemporal dynamic correlation calculation method based on gray relation analysis is proposed to construct dynamic adjacency matrices. Then, the spatiotemporal features are sufficiently extracted by the graph convolutional network and the multi-channel temporal convolutional network. Two real-world air quality datasets collected from Beijing and Fushun are applied to verify the performance of our proposed model. The experimental results show that compared with other baselines, the DGC-MTCN model has excellent prediction accuracy. Especially for the prediction of multi-step and different stations, our model performs better temporal stability and generalization ability.     

An algorithm for a decomposition of weighted digraphs: with applications to life cycle analysis in ecology

In the analysis of organism life cycles in ecology, comparisons of life cycles between species or between different types of life cycles within species are frequently conducted. In matrix population models, partitioning of the elasticity matrix is used to quantify the separate contributions of different life cycles to the population growth rate. Such partition is equivalent to a decomposition of the life cycle graph of the population. A graph theoretic spanning tree method to carry out the decomposition was formalized by Wardle [Ecology 79(7), 2539–2549 (1998)]. However there are difficulties in realizing a suitable decomposition for complex life histories using the spanning-tree method. One of the problems is the occurrence of life cycles that contain contradictory directions that defy biological interpretation. We propose an algorithmic approach for decomposing a directed, weighted graph. The graph is to be decomposed into two parts. The first part is a set of simple cycles that contain no contradictory directions and that consist of edges of equal weight. The second part of the decomposition is a subgraph in which no such simple cycles are obtainable. When applied to life cycle analysis in ecology, the proposed method will guarantee a complete decomposition of the life cycle graph into individual life cycles containing no contradictory directions. Although the research described in this article has been funded in part by the United States Environmental Protection Agency through STAR cooperative agreement R-82940201-0 to the University of Chicago, it has not been subjected to the Agency’s required peer and policy review and therefore does not necessarily reflect the views of the Agency and no official endorsement should be inferred.     

Kinetic modeling and fitting software for interconnected reaction schemes: VisKin

Reaction kinetics for complex, highly interconnected kinetic schemes are modeled using analytical solutions to a system of ordinary differential equations. The algorithm employs standard linear algebra methods that are implemented using MatLab functions in a Visual Basic interface. A graphical user interface for simple entry of reaction schemes facilitates comparison of a variety of reaction schemes. To ensure microscopic balance, graph theory algorithms are used to determine violations of thermodynamic cycle constraints. Analytical solutions based on linear differential equations result in fast comparisons of first order kinetic rates and amplitudes as a function of changing ligand concentrations. For analysis of higher order kinetics, we also implemented a solution using numerical integration. To determine rate constants from experimental data, fitting algorithms that adjust rate constants to fit the model to imported data were implemented using the Levenberg-Marquardt algorithm or using Broyden-Fletcher-Goldfarb-Shanno methods. We have included the ability to carry out global fitting of data sets obtained at varying ligand concentrations. These tools are combined in a single package, which we have dubbed VisKin, to guide and analyze kinetic experiments. The software is available online for use on PCs.     

乙、丁酸添加条件下丁醇发酵图论模型的构建

有机酸代谢途径在丁醇发酵过程中具有重要的作用,对细胞内碳流的分配和产物的合成影响显著。在7 L厌氧发酵罐中,进行了间歇添加乙酸或丁酸的发酵实验。结果表明,乙、丁酸的添加显著提高了总溶剂的生产效率,分别提高了47.1%和39.2%;此外,丁醇/丙酮比在添加丁酸的批次中提高了21.7%,在添加乙酸的批次中降低了16.2%;厌氧瓶中的发酵实验也证实了以上结果。有机酸代谢计算的结果表明,乙、丁酸的添加基本上阻断了相应有机酸闭环的吸收途径。基于相关报道和代谢计算结果,构建了针对乙、丁酸添加批次的图论模型,并利用该模型对不同发酵条件下的溶剂浓度和丁醇/丙酮比进行了计算。结果表明,该模型很好地预测了实验结果,合理地构建了乙、丁酸添加批次的信号传递线图。     

基于TransformerMGI的microRNA靶向基因预测

生物小分子microRNA可以对基因表达进行正向或负向调控,研究microRNA与基因之间的关系对于机体稳态的维持和疾病治疗都有着重要意义。利用深度学习方法对microRNA和基因靶向关系进行预测,提出了TransformerMGI模型。在特征工程阶段,针对生物序列潜在信息难以准确地提取这一问题,TransformerMGI模型分别采用了基于图卷积神经网络的GP-GCN方法和DNA2Vec模型对microRNA和基因数据的潜在信息进行提取,得到了二者的表征嵌入矩阵,在模型方面,TransformerMGI模型引入了幂归一化来改进经典的深度学习模型。利用microRNA和基因数据经过特征提取后得到两个表征矩阵,这两个矩阵分别被放入TransformerMGI模型中,通过TransformerMGI模型内部的Attention机制对二者自身和相互的特征信息进行了聚合和关联运算,最终预测出microRNA调控基因的概率。采用ROC曲线下面积和准确召回率曲线作为模型性能评价指标,将TransformerMGI与其他现有模型进行了比较评估。实验结果表明,TransformerMGI模型的AUC和AUPRC评分均可达0.91以上,优于现有的其他模型。TransformerMGI模型能在不考虑生物学原理和基因组背景的前提下,仅依赖microRNA和基因的碱基序列信息,实现microRNA靶向基因的预测,从而为后续的microRNA靶向基因预测研究提供了可借鉴的深度学习方法。     

关于生态场的几点评述

依据植物生态场的系统研究资料,对生态场的概念、场的特征函数、生态场的图形以及生态场对植物相互作用的解释进行了评述.阐明了生态场的最基本属性是物质性,目前的研究水平尚不能确定生态场是否是有别于一般物理场的特殊形式的场.对生态场的基本特征函数-生态势模型给予了形式与模型参数估测方面的评述,表明两种生态势模型各有一定的特点与优越性.作者强调,生态场更具生态学的方法论意义,应用生态场对植物相互作用形式与过程的分析描述,具有定量化、综合性与直观化特点.生态场的绘图为生态学研究开辟了新的途径,特别是应用生态场理论分析群体植物的动态变化过程(由互利共存到干涉竞争)会有合理的定量解释.