While quantitative structure activity relationship (QSAR) models have been employed successfully for the prediction of small model protein chromatographic behavior, there have been no reports to date on the use of this methodology for larger, more complex proteins. Recently our group generated a homologous library of antibody Fab fragments with different combinations of surface hydrophobicities and electrostatic potentials, and demonstrated that the unique selectivities of multimodal resins can be exploited to separate these Fab variants. In this work, results from linear salt gradient column experiments are used to develop QSAR models for the prediction of Fab retention in six chromatographic systems, including multimodal (Capto MMC, Nuvia cPrime, and two novel ligand prototypes), HIC (Capto Phenyl), and CEX (CM Sepharose Fast Flow) resins. These single resin models utilize newly developed “local descriptors” to quantify changes around point mutations in the Fab library and also include novel cluster based descriptors recently introduced by our group. Subsequent rounds of feature selection prune the initial descriptor list to a concise set of four to seven descriptors for each resin system. Several machine learning algorithms are used to generate validated models with high training set correlations (R<sup>2</sup>&gt;0.73) that are well suited for predicting salt elution concentrations. Notably, four out of the six models achieved this fit and predictive accuracy using a linear algorithm rather than a more complex non-linear method. These results represent the first successful utilization of QSAR for the prediction of the chromatographic behavior of complex proteins such as Fab fragments. The QSAR model development methodology presented here can be implemented to facilitate process development for purification of recombinant biopharmaceuticals from product-related impurities through <i>in silico </i>screening of chromatographic resin alternatives and opens up the possibility of developing such models for even more complex proteins.

QSAR models for prediction of chromatographic behavior of homologous Fab variants

251st National Meeting (2016)

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.