Neal Lab
Multivariate Optical Spectroscopy of
Biomolecular Assemblies in Complex Fluids
It is clear that solutes can have very different interactions in the heterogeneous microdomains of complex fluids, but unraveling the fluorescence response of solutes that are distributed over fluid microdomains requires specialized data analysis. Multivariate data acquisition facilitates analysis of such mixtures because multivariate data sets are subject to matrix and tensor analysis methods that separate multivariate signals into components without a priori knowledge of the number of components or the specifics of their spectral properties. Our goal in data analysis software development has been to generate methods that preserve this generality while maximizing the incorporation of the raw information in the data set into the separated component responses. Our work resolving multivariate measurements into component responses (spectra, etc) using multivariate statistics and constrained optimization demonstrates this approach. Currently we are investigating the application of powerful tools from related fields such as system identification, control theory and machine learning to multivariate fluorescence analysis in the development of methods that accommodate the non-ideal photokinetic behavior that is sometimes observed in complex fluids. 
