Cruz-Bournazou, M. N. | bioRxiv Science

bioRxiv2023

Automation of Experimental Workflows for High Throughput Robotic Cultivations

Process systems engineering methods and tools have been difficult to apply in bioprocess engineering, mainly due to the high complexity of biological systems and the low reproducibility of the experiments. High throughput robotic cultivation platforms in combination with computational tools for experimental design, resource scheduling, and operation, are rapidly gaining popularity. One important contribution being the generation of data in high throughput needed to overcome this lack of data with high information content and the worrying reproducibility crisis in life sciences. In this work, directed acyclic graphs are used to represent, manage and track all experimental workflows in a robotic platform. They support data provenance and enable traceability and reproducibility of workflows in robotic facilities. The experimental workflows are automated using Apache Airflow enabling to manage all necessary steps for fed-batch cultivations, including sampling, sample transport by a mobile robot, feed additions, data collection, storage in a SQL database and model fitting. The added value of this system is demonstrated in scale-down experiments, where E. coli BL21 (DE3), producing elastin like proteins, exhibits robustness towards glucose oscillations that mimic industrial cultivation conditions.

bioengineering↗

bioRxiv2021

Hybrid Gaussian Process Models for continuous time series in bolus fed-batch cultures

Hybrid modeling, meaning the integration of data-driven and knowledge-based methods, is quickly gaining popularity among many research fields, including bioprocess engineering and development. Recently, the data-driven part of hybrid methods have been largely extended with machine learning algorithms (e.g., artificial neural network, support vector regression), while the mechanistic part is typically using differential equations to describe the dynamics of the process based on its current state. In this work we present an alternative hybrid model formulation that merges the advantages of Gaussian Process State Space Models and the numerical approximation of differential equation systems through full discretization. The use of Gaussian Process Models to describe complex bioprocesses in batch, fed-batch, has been reported in several applications. Nevertheless, handling the dynamics of the states of the system, known to have a continuous time-dependent evolution governed by implicit dynamics, has proven to be a major challenge. Discretization of the process on the sampling steps is a source of several complications, as are: 1) not being able to handle multi-rate date sets, 2) the step-size of the derivative approximation is defined by the sampling frequency, and 3) a high sensitivity to sampling and addition errors. We present a coupling of polynomial regression with Gaussian Process Models as representation of the right-hand side of the ordinary differential equation system and demonstrate the advantages in a typical fed-batch cultivation for monoclonal antibody production.

bioengineering↗