Related Experiment Videos
Is there a rational method to purify proteins? From expert systems to proteomics
1Centre for Biochemical Engineering and Biotechnology, Department of Chemical Engineering, Millenium Institute for Advanced Studies in Cell Biology and Biotechnology, University of Chile, Beauchef 861, Santiago, Chile. juasenjo@ing.uchile.cl
Journal of Molecular Recognition : JMR
|May 12, 2004
Summary
This study introduces a computer-based expert system that uses proteomic characterization to optimize recombinant protein purification. It efficiently selects chromatographic steps, minimizing effort while maximizing purity for therapeutic applications.
Area of Science:
- Biochemistry
- Proteomics
- Chemical Engineering
Background:
- Recombinant protein purification demands multiple chromatographic steps for high purity.
- Selecting optimal purification strategies from various techniques (ion exchange, hydrophobic interaction, gel filtration, affinity) is challenging.
- Complex mixtures and unknown contaminants complicate achieving desired purity levels (e.g., 99.9%).
Purpose of the Study:
- To develop and validate a computational expert system for selecting efficient chromatographic purification strategies.
- To minimize the number of purification steps while maximizing the purity of recombinant proteins.
- To leverage initial proteomic characterization for informed purification design.
Main Methods:
- Proteomic characterization of the initial protein mixture.
- Development of two algorithms: one for selecting purification methods based on physicochemical properties, and another for predicting contaminant levels and purity.
- Implementation of these algorithms into a computer-based expert system.
- Experimental validation using protein mixtures and a Bacillus subtilis fermentation supernatant.
Main Results:
- The expert system successfully predicted and optimized purification strategies.
- Experimental validation confirmed the system's effectiveness in achieving high protein purity.
- The system demonstrated robustness to typical experimental errors (<10%) in property determination.
Conclusions:
- An expert system approach, integrating proteomic data, offers an efficient method for designing recombinant protein purification processes.
- This computational strategy can significantly reduce the complexity and improve the outcome of protein purification.
- Accurate physicochemical property data is crucial for the system's predictive power and robustness.