简介:Traditionally,SQLquerylanguageisusedtosearchthedataindatabases.However,itisinappropriateforend-users,sinceitiscomplexandhardtolearn.Itistheneedofend-user,searchingindatabaseswithkeywords,likeinwebsearchengines.Thispaperpresentsasurveyofworkonkeywordsearchindatabases.ItalsoincludesabriefintroductiontotheSEEKERsystemwhichhasbeendeveloped.
简介:Althoughtheproteinsequence-structuregapcontinuestoenlargeduetothedevelopmentofhigh-throughputsequencingtools,theproteinstructureuniversetendstobecompletewithoutproteinswithnovelstructuralfoldsdepositedintheproteindatabank(PDB)recently.Inthiswork,weidentifyaproteinstructuraldictionary(Frag-K)composedofasetofbackbonefragmentsrangingfrom4to20residuesasthestructural"keywords"thatcaneffectivelydistinguishbetweenmajorproteinfolds.Wefirstlyapplyrandomizedspectralclusteringandrandomforestalgorithmstoconstructrepresentativeandsensitiveproteinfragmentlibrariesfromalargescaleofhigh-quality,non-homologousproteinstructuresavailableinPDB.Weanalyzetheimpactsofclusteringcut-offsontheperformanceofthefragmenthbraries.Then,theFrag-KfragmentsareemployedasstructuralfeaturestoclassifyproteinstructuresinmajorproteinfoldsdefinedbySCOP(StructuralClassificationofProteins).OurresultsshowthatastructuraldictionarywithN4004-to20-residueFrag-KfragmentsiscapableofclassifyingmajorSCOPfoldswithhighaccuracy.
简介:FRUITFULRESULTSDuringtheirhall-a-monthvisittoTibet,theColorfUlCausegroupsignedsome40contracts,agreementsandlettersofintent.Th...
简介:在这份报纸,我们在Rn上为Hormander鈥檚拉普拉斯算符考虑主要特征值问题,并且我们为如此的主要特征值发现一个比较原则。我们也在整个Rn学习一个相关半线性的亚elliptic方程并且证明在一个合适的条件下面,我们无穷地有这个问题的许多积极答案。关键词主要特征值-Hormander鈥檚拉普拉斯算符-积极答案先生(2000)题目分类35J65-35H20这个工作被中国的钥匙973工程部分地支持