| Summary: | The ultracentrifuge has traditionally been used to obtain molecular weights, but recent developments have expanded the capabilities of this instrument. One of these developments was the ability to obtain the complete molecular weight distribution (MWD) from ultracentrifugal data through the use of a linear programming calculation developed by Scholte.
One objective of this research was to apply the MWD calculation procedure of Scholte to the study of protein systems. In preparation for this work, the ultracentrifuge was equipped with new UV optics and integrated with a data acquisition system coupled with a desktop computer to allow the routine calculation of molecular weights from ultra-centrifugal data of both homogeneous and heterogeneous systems. The combination of ultracentrifuge and data acquisition system was found to work well in spite of baseline problems associated with the optical system.
A new calculation procedure for obtaining the MWD was also investigated. Development of the frequency vs. molecular weight function was based on multiple regression rather than linear programming and had the advantage of providing statistical parameters to assess the accuracy of the fitting procedures. A FORTRAN program was developed and tested in regard to its capabilities and limitations, through the analysis of model systems. The multiple regression approach was found to be equivalent to the method of Scholte. The availability of the statistical parameters led to the inclusion in the MWD program of a simplex optimization routine that allowed the evaluation of the weight average molecular weights and the concentrations of individual components present in multicomponent systems. Using the simplex routine to search for the best fit solutions theoretically allowed a reduction in the number of rotor speeds required to obtain a distribution, thus eliminating one of the major restrictions of Scholte's method.
Attempts to verify experimentally the theoretical advantages of this new approach only provided an indication
that the method was viable. Due to the limitations of the available equipment, the MWD calculation could not be put to use in studying protein systems, although the potential for its use in such work was demonstrated. === Land and Food Systems, Faculty of === Graduate
|
|---|
