识别药物-靶标相互作用(DTI)是药物再利用和创新药物发现中不可或缺的关键步骤,目前已经有许多基于序列的计算方法被广泛应用于DTI预测;然而,在以往的基于序列的研究中,特征提取通常只关注序列本身,忽视了异构信息网络,如药物-药物相互...识别药物-靶标相互作用(DTI)是药物再利用和创新药物发现中不可或缺的关键步骤,目前已经有许多基于序列的计算方法被广泛应用于DTI预测;然而,在以往的基于序列的研究中,特征提取通常只关注序列本身,忽视了异构信息网络,如药物-药物相互作用网络、药物-靶标相互作用网络等。因此,提出一种基于序列和多视角网络进行DTI预测的新方法 SMN-DTI(prediction of Drug-Target Interactions based on Sequence and Multi-view Networks)。该方法使用变分自编码器(VAE)学习药物SMILES(Simplified Molecular-Input Line-Entry System)字符串和靶标氨基酸序列的嵌入矩阵;随后,利用具有两级注意力机制的异构图注意力网络(HAN)从节点和语义2个视角的网络中聚集来自药物或靶标的不同邻居的信息,并得到最终的嵌入。在2个广泛用于DTI预测的基准数据集Hetero-seq-A和Hetero-seqB上对SMN-DTI和基准方法进行评估的结果表明,在3种不同正负样本比例下SMN-DTI均取得了最优的特征曲线下面积(AUC)和精确召回曲线下面积(AUPR)。可见,SMN-DTI比目前主流的先进预测方法具有更好的性能。展开更多
变分图自编码器是图嵌入研究中重要的深度学习模型,但存在着先验正态分布缺陷、训练过程中容易出现后验塌陷等问题.本文从建立云概念空间与隐空间的映射关系入手,引入云模型数字特征对网络中的节点进行不确定性概念表示,设计了一种基于...变分图自编码器是图嵌入研究中重要的深度学习模型,但存在着先验正态分布缺陷、训练过程中容易出现后验塌陷等问题.本文从建立云概念空间与隐空间的映射关系入手,引入云模型数字特征对网络中的节点进行不确定性概念表示,设计了一种基于多维云模型的变分图自编码器(Variational Graph Autoencoder based on Multidimensional Cloud Model,MCM-VGAE).该模型实现了隐空间的多维云概念嵌入及相应的漂移性损失度量,将先验分布扩展为泛正态分布,利用多维正向云发生器及云包络带修正采样算法实现了重参数化过程,有效缓解了后验塌陷现象.在应用效果上,模型在多类型数据集上的链路预测、节点聚类、图嵌入可视化实验表现均优于基准模型,进一步说明了方法的普适有效性.展开更多
Stocks that are fundamentally connected with each other tend to move together.Considering such common trends is believed to benefit stock movement forecasting tasks.However,such signals are not trivial to model becaus...Stocks that are fundamentally connected with each other tend to move together.Considering such common trends is believed to benefit stock movement forecasting tasks.However,such signals are not trivial to model because the connections among stocks are not physically presented and need to be estimated from volatile data.Motivated by this observation,we propose a framework that incorporates the inter-connection of firms to forecast stock prices.To effectively utilize a large set of fundamental features,we further design a novel pipeline.First,we use variational autoencoder(VAE)to reduce the dimension of stock fundamental information and then cluster stocks into a graph structure(fundamentally clustering).Second,a hybrid model of graph convolutional network and long-short term memory network(GCN-LSTM)with an adjacency graph matrix(learnt from VAE)is proposed for graph-structured stock market forecasting.Experiments on minute-level U.S.stock market data demonstrate that our model effectively captures both spatial and temporal signals and achieves superior improvement over baseline methods.The proposed model is promising for other applications in which there is a possible but hidden spatial dependency to improve time-series prediction.展开更多
文摘识别药物-靶标相互作用(DTI)是药物再利用和创新药物发现中不可或缺的关键步骤,目前已经有许多基于序列的计算方法被广泛应用于DTI预测;然而,在以往的基于序列的研究中,特征提取通常只关注序列本身,忽视了异构信息网络,如药物-药物相互作用网络、药物-靶标相互作用网络等。因此,提出一种基于序列和多视角网络进行DTI预测的新方法 SMN-DTI(prediction of Drug-Target Interactions based on Sequence and Multi-view Networks)。该方法使用变分自编码器(VAE)学习药物SMILES(Simplified Molecular-Input Line-Entry System)字符串和靶标氨基酸序列的嵌入矩阵;随后,利用具有两级注意力机制的异构图注意力网络(HAN)从节点和语义2个视角的网络中聚集来自药物或靶标的不同邻居的信息,并得到最终的嵌入。在2个广泛用于DTI预测的基准数据集Hetero-seq-A和Hetero-seqB上对SMN-DTI和基准方法进行评估的结果表明,在3种不同正负样本比例下SMN-DTI均取得了最优的特征曲线下面积(AUC)和精确召回曲线下面积(AUPR)。可见,SMN-DTI比目前主流的先进预测方法具有更好的性能。
文摘变分图自编码器是图嵌入研究中重要的深度学习模型,但存在着先验正态分布缺陷、训练过程中容易出现后验塌陷等问题.本文从建立云概念空间与隐空间的映射关系入手,引入云模型数字特征对网络中的节点进行不确定性概念表示,设计了一种基于多维云模型的变分图自编码器(Variational Graph Autoencoder based on Multidimensional Cloud Model,MCM-VGAE).该模型实现了隐空间的多维云概念嵌入及相应的漂移性损失度量,将先验分布扩展为泛正态分布,利用多维正向云发生器及云包络带修正采样算法实现了重参数化过程,有效缓解了后验塌陷现象.在应用效果上,模型在多类型数据集上的链路预测、节点聚类、图嵌入可视化实验表现均优于基准模型,进一步说明了方法的普适有效性.
文摘Stocks that are fundamentally connected with each other tend to move together.Considering such common trends is believed to benefit stock movement forecasting tasks.However,such signals are not trivial to model because the connections among stocks are not physically presented and need to be estimated from volatile data.Motivated by this observation,we propose a framework that incorporates the inter-connection of firms to forecast stock prices.To effectively utilize a large set of fundamental features,we further design a novel pipeline.First,we use variational autoencoder(VAE)to reduce the dimension of stock fundamental information and then cluster stocks into a graph structure(fundamentally clustering).Second,a hybrid model of graph convolutional network and long-short term memory network(GCN-LSTM)with an adjacency graph matrix(learnt from VAE)is proposed for graph-structured stock market forecasting.Experiments on minute-level U.S.stock market data demonstrate that our model effectively captures both spatial and temporal signals and achieves superior improvement over baseline methods.The proposed model is promising for other applications in which there is a possible but hidden spatial dependency to improve time-series prediction.
