Model Selection: Developing a workflow for selection of optimal model from several possible computational models and their associated parameter estimation using a hybrid approach consisting of genetic algorithm and classical optimizer based on data present in literature.
Dynamic Optimization: Selecting the design parameters from the process variables and using them in a dynamic optimization framework to get optimal profiles for cell concentration, dissolved oxygen rate, pH of the system, glucose concentration, cell viability, cell density and protein concentration such that the objective functional(s) is/are extremized.
ANN based Optimal Control Reformulation: Development of ANN based algorithm for solving the OC problem and testing along with the benchmarks.
Data Acquisition Setup: Development of an experimental set up comprising of 3L bioreactor and a set of sensors that can be used for online monitoring of oxygen , pH, cell density and glucose level using labview system.
Validation of AI based Tool: A proof of concept for dynamic data acquisition from the bioreactor for recombinant protein production using baculovirus expression system and validation of the proposed AI based tool for dynamic optimization.
Construction of predictive models (unstructured) for baculovirus infection and optimal model selection based on cell density and cell viability data.
Parameter estimation for unstructured model using GA that can be used for prediction of virus/protein production through several passages.
Construction of a novel framework that enables AI based RNN to emulate unstructured model while determining the optimal structure of RNN balancing AI model accuracy and overfitting.
Rising over the criticism of AI based models use less physics involved in the process, PINN based hybrid approach amalgamating the best of AI as well as physics of the process has been utilized to build surrogate models which can behave like Physics based models but at the same time doesn’t face difficulties involved in numerical integration of stiff mathematical equations involved in the unstructured model.
Optimal control of bioreactor for finding the optimal recipes of substrate and virus addition considering uncertainty in the experiments using robust optimization.
Design and fabrication of the affordable biocompatible headplate setup used for the online data acquisition of protein expression using insect cells in semi-batch mode. This set up was also used for validation of the model prediction.
Comparing the cell viability, cell density, oxygen profile of the semi-batch process with and without control of dissolved oxygen using the 3D printed headplate reactor.
Implementation of transfer learning and deep learning (YOLO8 based object detection model) for the automated estimation of the insect cell size. This leads to a tool for online detection of infection level from bioreactor using a label-free approach and potentially eliminate the lengthy procedure of end point dilution.
Sharma, S., Mahadevan, J., Giri, L. and Mitra, K., 2023. Identification of optimal flow rate for culture media, cell density, and oxygen toward maximization of virus production in a fed‐batch baculovirus‐insect cell system. Biotechnology and Bioengineering, 2023, 120(12), pp.3529-3542.
Sharma, S., Keerthi, P.N., Giri, L. and Mitra, K., 2022. Toward Performance Improvement of a Baculovirus–Insect Cell System under Uncertain Environment: A Robust Multi-Objective Dynamic Optimization Approach for Semi-batch Suspension Culture. Industrial & Engineering Chemistry Research, 2023, 62(1), pp.111-125.
Singh, R., Sharma, S., Kareenhalli, V. V., Giri, L., Mitra, K., Experimental investigation into Indole production using Passaging of E. coli and B. subtilis along with unstructured modeling and parameter estimation using Dynamic Optimization: An Integrated Framework. Biochemical Engineering Journal, 2020, 163, 107743.
Mahadevan, J., Bagoria, N., Neelapala, S., Shrikanth, D., Jana, S., Mitra, K., Giri, L., Towards fluorescent-tag-Less Viral titration: Automated Estimation of cell-Size Distribution and Infection Level from phase-Contrast Microscopy Using Deep Learning and Transfer Learning, 46th annual IEEE/EMBS Conference, Orlando, USA, July 2024 (accepted).
Masampally, V.S., Sharma, S., Giri, L., Mitra, K., Physics Informed Neural networks for Baculovirus-Insect Cell System, 2023 Ninth Indian Control Conference (ICC), Visakhapatnam, India, Dec 2023, pp. 22-27, doi: 10.1109/ICC61519.2023.10442232.
Sharma, S., Giri, L. and Mitra, K., Multi-objective Optimization and control under Uncertainty for performance improvement of a Baculovirus Expression Vector System, 2022 Eighth Indian Control Conference (ICC), Madras, India, Dec 2022, pp. 416-421, doi: 10.1109/ICC56513.2022.10093623.
Sharma, S., Pujari, K.N., Miriyala, S.S., Giri, L. and Mitra, K., Unstructured modelling and RNN Surrogate development for optimizing vaccine production in Baculovirus Expression Vector System. 2021 Seventh Indian Control Conference (ICC), Mumbai, India, Dec 2021, pp. 406-411, doi: 10.1109/ICC54714.2021.9703114.
Miriyala, S. S. and Mitra, K., Optimal Control using Evolutionary Algorithms through Neural Network based TRANSFORMation, 2020 IEEE Symposium Series on Computational Intelligence (SSCI), Canberra, ACT, Australia, Dec 2020, pp. 1379-1386, doi: 10.1109/SSCI47803.2020.9308475.