Optimal synthesis of protein purification processes.

There has been an increasing interest in the development of systematic methods for the synthesis of purification steps for biotechnological products, which are often the most difficult and costly stages in a biochemical process. Chromatographic processes are extensively used in the purification of multicomponent biotechnological systems. One of the main challenges in the synthesis of purification processes is the appropriate selection and sequencing of chromatographic steps that are capable of producing the desired product at an acceptable cost and quality. This paper describes mathematical models and solution strategies based on mixed integer linear programming (MILP) for the synthesis of multistep purification processes. First, an optimization model is proposed that uses physicochemical data on a protein mixture, which contains the desired product, to select a sequence of operations with the minimum number of steps from a set of candidate chromatographic techniques that must achieve a specified purity level. Since several sequences that have the minimum number of steps may satisfy the purity level, it is possible to obtain the one that maximizes final purity. Then, a second model that may use the total number of steps obtained in the first model generates a solution with the maximum purity of the product. Whenever the sequence does not affect the final purity or more generally does not impact the objective function, alternative models that are of smaller size are developed for the optimal selection of steps. The models are tested in several examples, containing up to 13 contaminants and a set of 22 candidate high-resolution steps, generating sequences of six operations, and are compared to the current synthesis approaches.