The Research Of Spectrum Analysis In Plasma ~1H-NMR Metabolomics

Metabolomics is a new generation of omics techniques that combine closely with clinical research and developed rapidly in recent years.Metabolomics has been one of the research hotspot because of it could analysis the molecule metabolites of complicated biological samples[1].Nuclear Magnetic Resonance(NMR)and Mass spectra(MS)are the two most important analysis platform in metabolomics at present,but NMR has been widely used because of its simple sample preparation,high precision,nondestructive and non-deviation,all of these character make it have a great superiority in analyzing the metabolites of biological sample.In the head research of NMR metabolomics,In order to maximize the sample availability information,researcher have done many research in sample preparation,NMR experimental design,temperature,the setting of relaxation time,even the suppression method of solvent peak.Also for the NMR spectra,which have developed many processing method in order to convenient for the next spectra data processing,like correction of phase,baseline correction,spectra aligning,normalization processing,integration,deconvolution,quantitative method.But the biological samples such as plasma,urine and cerebrospinal fluid,there are so many molecular metabolites that caused NMR spectra a lot of overlap inevitable.Metabolite identification according to analyzing the NMR spectra is the key of NMR metabolomics next development,but the overlap of spectra peak is biggest challenge of NMR metabolomics.The traditional solution to solve peak overlap is expand the NMR experiment to two-dimentional homonuclear(such as JRES,COSY and TOCSY)or heterocaryosis(such as HSQC and HMBC),but the two-dimentional NMR experiment time is 10-40 times compare to 1H-NMR,therefore,the two-dimentional NMR is not suitable for high throughput and large sample of NMR metabolomics.When in the process of moving from trial to clinical practice,NMR metabolomics need to detect the metabolites change fast and exactly,so we need to explore how to solve the NMR spectra of mixture qualitative and quantitative.With the development of research,we found there are many metabolite database such as Human Metabolome Database(HMDB)and Madison Metabolomics Consortium Database(MMCD)have been build and supplied the method which can identify metabolites automatically according to chemical shift,it is tremendously convenient the metabolite identification of mixture NMR spectra.With the present of standard spectra database,there are some method that identify metabolite automatically has been appeared,but these present methods also have many deficiency.In order to verify the algorithm that we build and also for develop more accurate metabolite identification for the future,this research has collected 1H-NMR spectra chemical shift of HMDB metabolite and build our own 1H-NMR metabolite database based on MySQL.Then we combined Jaccard coefficient and match radio designed our match algorithm about chemical shift based on the match algorithm of HMDB and MMCD,amend the bias about multi-peak and less-peak of the present match algorithm,increased the precision and rationality of the metabolite identification.