Abstract: A predictive model is described that can predict parameter concentrations in the future based on initial, measured concentrations and historical data. A plurality of multivariate techniques can be used to construct the predictive model capable of forecasting concentrations over multiple and diverse cell lines. The predictive model is also scalable. In one embodiment, a future lactate concentration trajectory is determined and at least one condition within a bioreactor is changed or modified to maintain lactate concentration within desired ranges.

Abstract: A predictive model is described that can predict parameter concentrations in the future based on initial, measured concentrations and historical data. A plurality of multivariate techniques can be used to construct the predictive model capable of forecasting concentrations over multiple and diverse cell lines. The predictive model is also scalable. In one embodiment, a future lactate concentration trajectory is determined and at least one condition within a bioreactor is changed or modified to maintain lactate concentration within desired ranges.

Abstract: A process and system for efficiently producing reference data that can be fed into a predictive model for predicting quality attribute concentrations in cell culture processes. A perfusion bioreactor is operated at pseudo-steady-state conditions and one or more attribute influencing parameters are manipulated and changed over time. As the one or more attribute influencing parameters are manipulated, one or more quality attributes are monitored and measured. In one embodiment, multiple quality attributes are monitored and measured in parallel. The quality attribute information is recorded in conjunction with the changes in the attribute influencing parameters. This information is then fed to the predictive model for propagating cell cultures in commercial processes and maintaining the cell cultures within desired preset limits.

Abstract: A predictive model is described that can predict parameter concentrations in the future based on initial, measured concentrations and historical data. A plurality of multivariate techniques can be used to construct the predictive model capable of forecasting concentrations over multiple and diverse cell lines. The predictive model is also scalable. In one embodiment, a future lactate concentration trajectory is determined and at least one condition within a bioreactor is changed or modified to maintain lactate concentration within desired ranges.