How is data science revolutionizing catalyst discovery and development to meet global energy, sustainability, and healthcare needs? In their recently published review in Nature Catalysis, Manu Suvarna and Javier Pérez-Ramírez from ETH Zürich and NCCR Catalysis assess the use of data-driven strategies in heterogeneous, homogeneous, and enzymatic catalysis. By analyzing catalytic tasks, model reactions, data representations, and machine learning algorithms, they highlight key advancements and opportunities for knowledge transfer within the field. The authors also identify a significant gap in the application of data science to experimental catalysis and propose greater integration of data-driven tools and concepts in experimental workflows.
“Data science and machine learning offer transformative approaches in catalysis research by complementing existing catalyst design toolboxes. These methods can accelerate the discovery and optimization of catalytic systems and unveil hidden patterns in data that map structure-property-performance relationships. However, realizing the full potential of digital catalysis requires fostering a collaborative mindset between the data science and catalysis communities”, says Manu, a data scientist working in an experimental catalysis group.
Emphasizing the need for data standardization, and by elaborating on four pillars of data science: descriptive, predictive, causal and prescriptive analytics with examples from catalysis, Manu and Javier’s article serves as a guide for catalysis practitioners and data scientists alike 🧭🧑🏽🔬 and aims to drive future research in digital catalysis 💻🧪.
🔸 Read the full publication here: https://lnkd.in/dyHtve9j
🔸 The article is featured on the Nature Catalysis cover: https://lnkd.in/dcUNnY2N
#SustainableChemistry #DigitalChemistry #Catalysis #DataScience #Research #RecommendedRead
#Cheminformatics #HeterogeneousCatalysis #HomogeneousCatalysis #Biocatalysis
Swiss National Science Foundation SNSF