摘要：In view of huge search space in drug design, machine learning has become a powerful method to predict the affinity between small molecular drug and targeting protein with the development of artificial intelligence technology. However, various machine learning algorithms including massive different parameters make the prediction framework choice to be quite difficult. In this work, we took a recent drug design competition(from XtalPi company on the DataCastle platform) as the typical case to find the optimized parameters for different machines learning algorithms and the most effective algorithm. After the parameter optimizations, we compared the typical machine learning methods as decision tree(XGBoost, LightGBM) and artificial neural network(MLP, CNN) with root-mean-square error(RMSE) and coefficient of determination(R^2) evaluation. As a result, decision tree is more effective than the neural network as LightGBM>XGBoost>CNN>MLP in the affinity prediction of the specific drug design problem with ～160000 samples. For a much larger screening task in a more complicated drug design study, the sophisticated neural network model may go beyond the decision tree algorithm after generalization enhancing and overfitting reducing. The advanced machine learning methods could extract more information of protein-ligand bindings than traditional ones and improve the screen efficiency of drug design up to 200–1000 times.

摘要：The effect of an external electric field on the bandgap is observed for two proposed heterostructures graphitic carbon nitride-graphene-hexagonal boron nitride(g-C_3N_4/G/h-BN) in hexagonal stack(AAA) and graphene-graphitic carbon nitridehexagonal boron nitride(G/g-C_3N_4/h-BN) in Bernal stack(ABA). Their inter-layer distance, binding energy and effective mass has also been calculated. The structure optimization has been done by density functional theory(DFT) with van der Waals corrections. The inter-layer distance, bandgap, binding energy and effective mass has been listed for these heterostructures and compared with that of bilayer graphene(BLG), graphene-hexagonal boron nitride(G/h-BN) hetero-bilayer, graphene-graphitic carbon nitride(G/g-C_3N_4) hetero-bilayer and graphitic carbon nitride-graphene-graphitic carbon nitride(g-C_3N_4/G/g-C_3N_4) heterostructure in Bernal and hexagonal stack. g-C_3N_4/G/h-BN is found to offer lower effective mass and larger bandgap opening among the considered heterostructures.

摘要：针对如何利用高性能多核化设备,提高网络安全产品的处理能力,设计和实现了一种基于x86架构的Llinux平台多核绑定技术。该技术首先建立DMA缓冲队列映射,减少网卡访问次数,采用SIMD多核思想设计和实现了虚拟数据桶,并对进入数据桶的数据实施负载均衡;将Netfilter主函数多线程化,并采用内核线程绑定技术将多线程绑定到指定核.实验结果表明,DMA缓冲队列映射可以提高网络设备的I/O吞吐量,虚拟数据桶减少了数据包二次拷贝的开销,节省内核态内存,多核绑定技术提高网络安全设备多核利用率和数据包处理能力。

摘要：针对EMIF规范中XML报文处理,基于Delphi及XML Data binding技术设计并实现了XML报文的生成与解析方法,为参照EMIF规范进行教育管理信息系统开发或利用Delphi进行XML相关开发的人员提供参考。

摘要：蛋白质的研究在如今的科学研究中的地位越来越重要,尤其是当今科技发展速度之迅速,人们所了解的蛋白质信息越来越多,也更加繁杂。而如何在这纷繁的信息之海中高效与精确地搜寻到所需要的蛋白质信息,是需要研究的问题。本文以NCBI和binding DB为例,设计一种蛋白质搜索信息整合方法,通过从搜索到的信息条目中提取关键字组成二元组并进行分组,在每个分组里同样进行细化关键字的提取和分组,以此循环,并且由二元组衍生到N元组,从而达到去除冗余信息和信息排序整合的目的。