EuOpstart - Big Data of Protocells: A Revolution in the Chemical Sciences

The high complexity and information-rich hyper-dimensionality of many chemical systems of technological relevance demands from chemists of the XXI century to get on the wave of big data and artificial intelligence (AI). Machine learning (ML), a subfield of AI lying on statistical algorithms that learn with training, will have a great impact on the generation, testing and refining of chemical models. Because ML techniques are suitable to address complex problems involving massive combinatorial spaces and non- linear behaviors, they are ideal to address one of the major scientific and technological current challenges, that is the long-term dream of implementing life-like artificial constructs. Although the ultimate goal of this research agenda, the bottom-up construction of living protocells, is still far from achievable, a new generation of chemists endowed with a systems view is starting to realize the enormous potential of being able to synthesize, analyze and exploit complex molecular networks and ensembles.
The main objective of the present project is then to interface the systems chemistry insight into the material world with computer AI, by studying big data from complex recursively-generated functional molecular networks through ML methods, as an indispensable hybrid technology to pave the way in protocell research. To do so, we propose to merge three different and compatible methodological approaches: (i) protocell research and reaction networks, assisted by highthroughput microfluidics processing, (ii) metabolomics-based analysis, and (iii) big data / machine learning. Importantly, the resulting interface technology will find application in multiple industrial processes where complex sets of reactions, interconnected in space and/or time, are involved. Some examples include drug discovery, chemical synthesis planning, bio-refinery, food technology and environmental chemistry.