We gather available genomic data for the cell type of interest. (Genomes of most commonly cultured cells have been sequenced.)
Based on its genome, we build a metabolic model of the cell. It includes all its metabolites, the chemical reactions implying them and their transfers in and out of the cell.
With our specific design of experiment, a limited number of wet lab experiment are carried.
We then combine data obtained through these experiments with literature data and expert knowledge to create a machine learning model that capture relevant aspect of the cell culture.
We use the AI model, specific to your cell line, to inform and constraint the metabolic model obtained on step 2. We then use our optimization algorithm to find the media, feed compositions and feeding strategy that suits your cells the most. This optimization can be done for any chosen property or a combination of them (productivity, growth, …).