analysi equin protein dcode gene structur express rna dsequenc probabilist framework align pair dend rna dseq data function neighbor infer relationship nonhomolog protein famili famili dspecif pack motif privaci preserv social network sensit edg weight approxim frequent itemset frequent itemset mine popular import step analyz data set broad rang applic tradit exact approach find frequent itemset requir item itemset occur support transact real data typic subject nois presenc nois tradit itemset mine fail detect relev itemset larg itemset vulner nois paper propos approxim frequent itemset afi nois toler itemset model addit usual requir suffici support transact afi model place constraint fraction error permit item column fraction error permit support transact constraint window approxim itemset exhibit systemat error context simpl nois model demonstr afi recov underli data pattern identifi fewer spuriou pattern exact frequent itemset approach exist error toler itemset approach yang al bay estim phylogenet reconstruct function neighbor infer relationship dhomolog protein famili famili dspecif pack motif describ approach infer function relationship homolog protein famili statist enrich altern function predict classif hierarchi gene ontolog structur classif protein scop protein structur repres robust graph fast frequent subgraph mine algorithm appli protein famili gener set famili specif pack motif amino acid residu pack pattern share famili member infrequ protein function protein infer identifi motif characterist famili emploi familyspecif motif elucid function relationship famili scop hierarchi specif postul famili function relat famili statist enrich motif characterist famili number protein famili motif famili greater expect chanc function infer method annot protein unknown function establish function neighbor exist famili altern function protein approxim cluster distribut data stream abstract investig problem cluster distribut data stream median cluster stream data arriv distribut site commun rout tree distribut cluster high speed data stream challeng task due limit commun capac storag space comput power site paper propos suit algorithm comput approxim median cluster distribut data stream topolog set topologyoblivi height awar path awar algorithm reduc maximum node transmiss polylog oppos transmit raw data simul algorithm distribut stream system real synthet dataset compos million data practic algorithm reduc data transmiss small fraction origin data result algorithm scalabl respect data volum approxim factor number site mine approxim order preserv cluster presenc nois abstract subspac cluster attract great attent due capabl find salient pattern high dimension data order preserv subspac cluster proven import high throughput gene express analysi function relat gene express set experiment condit express pattern repres consist order attribut exist order preserv cluster model requir object cluster ident attribut order deviat real data noisi due measur technolog limit experiment variabl prohibit strict model reveal true cluster corrupt nois paper studi problem reveal order preserv cluster presenc nois propos nois toler model call approxim order preserv cluster aopc requir object cluster ident attribut order requir fraction object ident attribut order object cluster deviat consensu order fraction attribut propos algorithm mine aopc experi gene express data demonstr effici effect algorithm increment subspac cluster multipl data stream poclust lossless cluster dissimilar data poclust lossless cluster dissimilar data cluster pair dwise dissimilar data partial order set ut unbalanc tree search benchmark abstract paper present unbalanc tree search ut benchmark design evalu perform eas program parallel applic requir dynam load balanc describ algorithm build varieti unbalanc search tree simul form load imbal creat version ut parallel languag openmp unifi parallel upc work steal mechan reduc load imbal benchmark perform ut parallel architectur includ sharedmemori system pc cluster found simpl implement ut upc openmp due upc share memori abstract result show upc openmp support effici dynam load balanc share memori architectur upc allevi underli commun cost distribut memori system dynam load balanc requir intens commun perform portabl remain difficult applic ut perform degrad pc cluster vari kei work steal paramet expos import tradeoff granular load balanc degre parallel commun cost mine approxim frequent itemset presenc nois algorithm analysi frequent itemset mine popular import step analysi data aris broad rang applic tradit exact model frequent itemset requir item occur support transact real data typic subject nois measur error date effect nois exact frequent pattern mine algorithm address primarili simul studi limit attent develop nois toler algorithm paper propos nois toler itemset model call approxim frequent itemset afi frequent itemset afi model requir itemset minimum number support transact mine approxim frequent itemset noisi data frequent itemset mine popular import step analyz data set broad rang applic tradit exact approach find frequent itemset requir item itemset occur support transact real data typic subject nois presenc nois tradit itemset mine fail detect relev itemset larg itemset vulner nois paper propos approxim frequent itemset afi nois toler itemset model addit usual requir suffici support transact afi model place constraint fraction error permit item column fraction error permit support transact constraint window approxim itemset exhibit systemat error context simpl nois model demonstr afi recov underli data pattern identifi fewer spuriou pattern exact frequent itemset approach exist error toler itemset approach yang al biclust gene express data tendenc advent dna microarrai technolog revolution experiment studi gene express cluster popular approach analyz gene express data proven success applic work focus discov subset gene exhibit similar express pattern subset condit gene express matrix specif order preserv cluster op cluster subset gene induc similar linear order subset condit pioneer work opsm model enforc strict order share gene cluster includ model special case model robust opsm similarli express condit allow form order equival group restrict order group guid model design implement determinist algorithm opc tree discov op cluster experiment studi real dataset demonstr effect algorithm applic tissu classif cell cycl identif addit larg percentag op cluster exhibit signific enrich function categori impli op cluster carri signific biolog relev gene ontolog friendli biclust express profil sound cluster analysi gene express profil gene function predict base hypothesi gene similar express profil impli strong correl function biolog activ gene ontolog accept standard organ gene function categori gene function categori sophist relationship part overlap cluster algorithm gener gene cluster relationship natur reflect gene function categori hierarchi failur resembl relationship reduc confid cluster gene function predict paper present cluster reveal true subspac cluster high dimens subspac cluster approach discov meaning cluster high dimension space cluster high dimension space transcrib multipl distinct maxim cluster project subspac direct consequ cluster independ subspac overwhelmingli larg set overlap cluster significantli similar reveal true underli cluster propos similar measur overlap cluster adopt model gaussian tail hyper rectangl captur distribut subspac cluster set experi synthet dataset demonstr effect approach applic real gene express data reveal impress meta cluster expect biologist framework ontolog ddriven subspac cluster tradit cluster descript task seek identifi homogen group object base valu attribut domain knowledg justifi cluster cluster algorithm domain knowledg consider paper domain knowledg repres hierarch ontolog develop framework directli incorpor domain knowledg cluster process yield set cluster strong ontolog implic cluster process ontolog inform util effici prune exponenti search space subspac cluster algorithm algorithm gener automat interpret cluster result map natur hierarch organ subspac cluster signific categor enrich ontolog hierarchi experi set gene express data gene ontolog demonstr prune techniqu driven ontolog significantli improv cluster perform minim degrad cluster qualiti hierarch organ gene cluster hierarchi gene ontolog successfulli captur op dcluster cluster tendenc high dimension space cluster process group set object class similar object unknown hidden pattern data set definit similar subtl recent similar measur typic base distanc euclidean distanc cosin distanc paper propos flexibl power cluster model op cluster order preserv cluster model object similar subset dimens valu object induc rel order dimens cluster aris express level coregul gene rise fall synchron respons sequenc environ stimuli discoveri op cluster essenti reveal signific gene regulatori network determinist algorithm design implement discov signific op cluster set extens experi real biolog data set demonstr effect effici detect regul pattern 