Two sessions of 2 workshops will be held during the afternoon of Tuesday, October 27. Please register during the online registration process. Each workshop is available at a fee of 85€.
|Session 1||Workshop 1||Workshop 2|
Chairman: Dr Olivier Dapremont
Chairman: Dr Alessandro Butte
|14:00-15:30||Continuous Chromatography for Small Molecules: Principle, Process Development and Large Scale Manufacturing||
Introduction to Machine-Learning Methods for Process Development. From Multivariate Analysis to Hybrid Models
|Session 2||Workshop 3||Workshop 4|
Chairman: Dr Eric Francotte
Chairman: Prof. Massimo Morbidelli
Prep Chiral and Achiral SFC:
|Application of Machine-Learning Methods to the Production of Therapeutic Proteins|
Workshop 1 - Continuous Chromatography for Small Molecules : Principle, Process Development and Large Scale Manufacturing
Dr Olivier Dapremont (AMPAC Fine Chemicals)
In this workshop, the students will learn the basic principles of continuous chromatography for the separation of small chiral molecules using the Simulated Moving Bed process. The instructor will introduce the mathematical modelling using the triangle theory to evaluate the performance of the separation and the influence of various parameters. Once the basic concepts are in place, the instructor will discuss improvements to the technology and the extension of the process to more complex separations. Finally, the student will learn how to develop an SMB separation from the initial screening to the implementation at commercial scale.
Dr Dapremont received his Ph.D. degree in Chemical Engineering and Applied Chemistry in Simulated Moving Bed technology (SMB) and chiral applications in 1997 from the University of Pierre and Marie Curie in Paris France.
He started his career, in 1992, developing SMB technology for Prochrom R&D, France. In 1997, He joined Chiral Technologies Europe, France, were he was in charge of the kilo lab for SMB chiral separations service.
He joined Aerojet Fine Chemicals, now AMPAC Fine Chemicals (AFC) in Rancho Cordova, CA, at the beginning of 2001. At AFC, Dr. Dapremont is in charge of the development of continuous processes for APIs and intermediates including chromatographic processes as well as implementing flow chemistry and related techniques at large scale. To this day, he has developed and implemented over 50 chiral and non-chiral separations using SMB from kilogram to multi ton scale.
Dr. Dapremont is author and co-author of several articles on preparative chromatography and SMB applications in various scientific journals and magazines. He is co-inventor on multiple patents using SMB as a purification step for APIs and he is a recognized expert in the field for the past 25 years.
Dr. Dapremont is also a member of the Organizing Committee of the Prep Symposium conference and a member of the Scientific Committee of the SPICA conference.
Workshop 2 - Introduction to Machine-Learning Methods for Process Development. From Multivariate Analysis to Hybrid Models
Dr Alessandro Butte (ETH Zürich)
In the last years, the use of machine learning (ML) algorithms has increase exponentially also in the context of process development and manufacturing. These tools, together with the progressive digitalization of all processes, procedures and data, the use of advanced sensors, and the possibility of putting different process units in direct communication with each other, is at the basis of what it has been called Industry 4.0.
This course is intended to introduce the participants to the main statistical and machine learning techniques that are typically used in the context of process development and manufacturing, with a particular view on the application to the so-called downstream processes and to chromatography in particular. The participants will be given an overview on all techniques, how and when to apply them, some theoretical background, and addition to relevant industrial application examples.
First, some basic statistical concepts (likelihood, p-values, etc.) will be given, together on a discussion on how to organize and prepare data. Then, an overview of the main multivariate data analysis (MVDA) methods will be given. This includes principle component analysis (PCA) and partial least square (PLS) regression. Then, an overview on the main ML methods for regression and classification will be given. This includes decision trees (DT) and random forests (RF), support vector machine (SVM), artificial neural networks (ANN), kernel ridge regression (KRR) and gaussian processes (GP). For all these techniques, focus will be put on the training and validation methods.
After introducing ML, the focus will move from purely data driven models to so-called hybrid models, where both mechanistic knowledge and data information are fused together to enhance data utilization and model robustness and predictability. Examples from chromatography will be given.
Finally, a brief overview will be given to experimental design and, in particular, on model based experimental design using the aforementioned methods.
Workshop 3 - Prep Chiral and Achiral SFC : Principles and Applications
Dr Eric Francotte (FrancotteConsulting GmbH, formerly, Executive Director at Novartis Pharma Research)
Workshop 4 - Application of Machine-Learning Methods to the Production of Therapeutic Proteins
Prof. Massimo Morbidelli (ETH Zürich)
In the last years, there has been an increased interest in continuous and integrated manufacturing of biopharmaceuticals, due to increasing cost and time pressure in industry, as well as new drug formats asking for more flexible production concepts. The next step towards intensification and modernization of current manufacturing technologies will be based on the use of online process control and optimization techniques. This requires the appropriate use of statistical approaches and machine learning techniques - not only in a continuous manufacturing scenario, but also within the more traditional batch integrated processes.
In this workshop, we discuss results obtained through hybrid approaches, where such machine learning techniques are used online in combination with traditional mechanistic models. Applications include : development of soft sensors for up and down stream monitoring (Raman based), online maintenance (prediction of protein A substitution in a monoclonal antibody capture process), hierarchical online process control and optimization of a chiral SMB process and a capture twin column unit.
The illustrated results mainly, but not only, refer to a end-to-end integrated continuos unit for the production of an industrial monovalent antibody of industrial relevance. They provide a very important basis to intensify the main advantages of continuous integrated manufacturing in agreement with the trends of industry 4.0.
Massimo Morbidelli received his Laurea in Chemical Engineering at the Politecnico di Milano and PhD at the University of Notre Dame. He is currently Professor at the Politecnico di Milano (Italy) and Professor Emeritus at the ETH Zurich (Switzerland).
His main research interest is in the area of biopharmaceutics and specifically on the integrated continuous manufacturing of therapeutic proteins, and its automation and digitalization.
Massimo Morbidelli is co-author of more than 750 papers, 23 international patents and six books, including the very recent ones on Continuous Biopharmaceutical Processes (2018), Cambridge University Press, coauthored with D Pfister and L Nicoud and Perfusion Cell Culture Processes for Biopharmaceuticals (2020), Cambridge University Press, coauthored with M Wolf and J-M Bielser. He is the first chemical engineer elected to the Italian Academy of Science (Accademia dei Lincei), serves as the Executive Editor of the ACS journal of Industrial & Engineering Chemistry Research, and is the recipient of:
In his career he advised more than 100 PhD students. He is a cofounder of ChromaCon Ltd., a spin-off company from his research group, which brings new chromatographic processes (MCSGP-technology) for the purification of proteins and peptides to the market (now acquired by YMC, Japan) and of DataHow Ltd. for the application of data science and machine learning in Biotechnology and specifically in the Biopharma Industry.