A High-Throughput Macromolecule Characterization System

Abstract

The size and complexity in structure of biopharmaceutical products makes them more susceptible to chemical or structural changes leading to lower potency or altered immunogenicity. Sustaining the stability of macromolecules becomes one of the greatest challenges in the development of biopharmaceutical products. The biophysical characterization of macromolecules is an essential step in stable formulation development. Structural changes of macromolecules in response to various environmental stresses or solution additives are measured using various techniques, and can then be analyzed using the empirical phase diagram (EPD). The empirical phase diagram (EPD) is a colored representation of overall structural integrity and conformational stability of macromolecules in response to various environmental perturbations. Numerous proteins and macromolecular complexes have been analyzed by EPDs to summarize results from large data sets from multiple biophysical techniques. The current EPD method suffers from a number of deficiencies including lack of a meaningful relationship between color and actual molecular features, difficulties in identifying contributions from individual techniques, and a limited ability to be interpreted by color blind individuals. Three improved data visualization approaches are proposed as techniques complementary to the EPD. Experimental data sets can be visualized as (1) RGB colors using three-index empirical phase diagrams, (2) equiangular polygons using radar charts, and (3) human facial features using Chernoff face diagrams. Recent development of high-throughput and multimodal spectrophotometers help rapidly collect the large volume of data that is required to create EPDs. Incompatible data formats of various instruments and heterogeneous analysis software are, however, standing in the way of quickly organizing and analyzing such large volumes of data. It is essential to develop dedicated analysis software for such biophysical data to achieve high-throughput systems, in terms of both hardware and software, for biophysical characterization of macromolecules. For this purpose, a web-based software framework called MiddaughSuite was developed in this work. The software was designed to easily handle data from various instruments, quickly analyze data using multiple mathematical functions, visualize data in the forms of graphs and diagrams including EPDs, radar chars and Chernoff face diagrams, and share data with other researchers.