Tissue Specific Significant Test And Functional Analysis For Promoter Sequence Motifs In Human Genes

Identification and analysis of tissue specific motif in non-coding area play animportant role in understanding mechanisms of the organism, disease diagnosis anddrug design. Discovering and analyzing these sequence motifs need reliable motifdiscovery algorithm, steady motif validation theory and systemic biological functionanalysis tools. This paper do research on sequences of human tissue specific genesand construct a motif discovery structure integrating many motif discovery algorithms.It also designs a scoring system for tissue specific motif basing on three statisticalhypothesis test methods and analyses functions of motif whose score is high.The contributions of this paper include the followings:Firstly, the recent motif discovery algorithms didn’t consider known prioriknowledge and result in over-estimate or low-estimate. My paper provide a structurefusing many existing motif discovery algorithm against this problem, it can merge allresults. According to the principle that Average is superior to single, our structurecan fuse all the advantages of other algorithms and make the result best. We discover3240motifs in3954genes of83tissues.Secondly, to examine statistical significance of sequence motif, we use two motifmatching score mechanism and three statistical hypothesis test methods to evaluatethe statistical significance of sequence motif. And we design a score mechanismmerging above-mentioned six combined evaluation results to grade sequence motif.We first define tissue rich motif (TRM) and tissue even motif (TEM). We discover673TRMs and917TEMs in3954genes of83human tissues.Finally, to map discovered motifs’ biological functions and find the evidence thatmotifs regulate biological genes’ functions, we design a systemic process frameworkanalyzing motifs’ functions and mapping functions to tissues. This framework canprovide more reliable evidence about result. In the experiment, most motifs find theirmapping to tissue functions and there are some tissues whose functions have goodresults when mapping to motifs. For example, for Fetallung tissue,14of16motifshave successful mapping. And there are6motifs whose functions all match itsresource tissue.