Premium
The Rational Selection of Purification Processes for Proteins: An Expert System for Downstream Processing Design a
Author(s) -
LESER E. W.,
ASENJO J. A.
Publication year - 1994
Publication title -
annals of the new york academy of sciences
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.712
H-Index - 248
eISSN - 1749-6632
pISSN - 0077-8923
DOI - 10.1111/j.1749-6632.1994.tb47406.x
Subject(s) - computer science , selection (genetic algorithm) , backup , process (computing) , resolution (logic) , expert system , data mining , database , machine learning , artificial intelligence , programming language
This paper describes the present state of the art in the use of ES for the rational selection of optimal purification processes for proteins (process synthesis). The main issues involved in selection of operations and process synthesis are reviewed. The high-resolution purification stages (one, two, three, or even more chromatographic steps where several alternatives in different orders of combinations can be used) have a combinatorial nature that can be represented by a "tree" structure. It can be analyzed by a depth-first or a breath-first approach. The two previous prototypes that have been developed are reviewed. The first prototype was exclusively based on "expert" rules, and its main weakness was in the selection of high-resolution purification operations for which a database of physicochemical properties of the main protein contaminants is needed. This was implemented in the second prototype in which a program for the calculation of separation coefficients on selection of high-resolution purification operations based on a preliminary database was interfaced. The present prototype being created uses a newly developed shell, which has an open architecture and uses object-oriented programming. This prototype uses a more "realistic" and accurate database (e.g., titration curves for contaminant proteins) and a tested method for selection of purification operations. The ES for selection of optimal protein separation sequences will give the user a number of alternatives chosen on the basis of the extensive data backup on proteins and unit operations.