AUTOMATED EXTRACTION OF PRINCIPAL COMPONENTS OF NON-STRUCTURAL PROTEIN 1 FROM SERS SPECTRUM

摘要

In this paper, the SERS analysis technique forextracting principal components of non-structural protein1 from its spectra is examined. The non-structural protein1 is found a major role in the replication process of virusof flaviviridae, the cause for many viral diseases. SERS isa technique that can provide fingerprint spectralinformation of even a single molecule. However, theRaman spectra from SERS complicate the featureextraction process with redundant features. PrincipalComponent Analysis is a signal processing technique,useful for filtering for the significant features whilefiltering off the redundant ones with minimal loss ofinformation. Here, PCA adopting a 3-steps approach, i.e.Eigenvalue-One-Criterion, Scree test and CumulativePercent Variance, is used to select significant principalcomponents of NS1 from Raman spectra. It is found thatprincipal components of NS1 from its spectra of [900x10]from SERS is found being trimmed to [9x10] by Screetest supplemented by EOC and [2x10] by CPV, with acorresponding reduction of 99% and 99.8% from theoriginal spectral array. However, since the spectra ofbiological samples in actual is noisier, selection by theformer of first nine components is found appropriate. Sofar, SERS analysis technique for detection of salivaryNS1 has yet to be reported.