PROT-EX : AN EXPERT SYSTEM FOR SELECTING THE SEQUENCE OF PROCESSES FOR THE DOWNSTREAM PURIFICATION OF PROTEINS

AUTOR(ES)

Eduardo Walter Leser

FONTE

IBICT - Instituto Brasileiro de Informação em Ciência e Tecnologia

DATA DE PUBLICAÇÃO

199607

RESUMO

This work presents the creation of an expert system prototype that suggests a sequence of operations for the downstream processing of proteins. A knowledge base with around six hundred logical rules was developed, structured to emulate the human reasoning. The process is divided into two parts: recovery, comprising the operations used from the end of fermentation to the production of a solution ready to be purified, and purification, the set of high resolution processes used to achieve the desired level of purity. For recovery the rules are based on relatively well established knowledge and built in a way that reproduces the operation in large scale production. For purification a rationale for the selection of a sequence of processes was developed. It consists of a system that compares extensive data on the product to be purified and data on the main contaminant proteins found in the different expression system used for biosynthesis. Parameters that translate the rules-of-thumb employed to guide the selection are defined. The deviation factor is a measure of the difference between individual properties of the product and the corresponding properties of each of the protein \"contaminants\". \ Efficiency\ measures how good a process is when exploiting the differences amongst properties. A concentration factor is included in the analysis. These variables define the selection separation coefficient and the comparison of its value determines the best possible method, based on this heuristic analysis, for the high resolution purification step. Data on the contaminants are obtained from databases that can be easily accessed by the programme. After each separation step there is a reduction in the amount of contaminant proteins. The number of steps must be sufficient to achieve a defined level of purity. The determination of the amount of \"contaminants\" eliminated is carried out using an algorithm developed for this purpose, based on simplified interpretations of chromatograms, that indicates the new concentration of the product after each step. The comparison of the concentration of the product after the separation with the defined level of purity indicates whether an additional step is necessary. The system can select among three chromatographic processes: size exclusion, hydrophobic interaction and ion-exchange. However, if the range of possible operations of ion-exchange at each pH is considered, the actual number of alternatives is twenty two. Affinity chromatography is included as an option given to the user during the selection. The implementation of the knowledge base was carried out in a commercial programme, Nexpert Object version 2.0, from Neuron Data Inc. Renowned industrial experts were consulted to clarify specific points for which knowledge is missing or ambiguous information is found when looking for industrial applications.

ASSUNTO(S)

biotecnologia purification of proteins bioquimica

Documentos Relacionados

• An expert system for performance-based direct delivery of published clinical evidence.
• Selecting a Computer System for the Clinical Laboratory
• Machine learning for an expert system to predict preterm birth risk.
• MBIS--an integrated system for the retrieval and analyses of sequence data from nucleic acids and proteins.
• Expanding the TRANSFAC database towards an expert system of regulatory molecular mechanisms.