The Chemical Knowledge Graph

Dr. Anirban Mudi has a Ph D in Chemistry. He is a science writer based in Bangalore, India. During his free time he synthesizes softwares for scientists. As luck would have it, I came across Bartosz Grzybowski et al’s triad of back-to-back papers in Angewandte Chemie (August 6, 2012) on applications of Chematica – a network of millions of chemical compounds and chemical reactions.  The name Chematica was inspired by Mathematica – a popular software used in scientific computing. Chematica caters to only organic synthesis as network analysis of inorganic compounds is difficult in their current chemical representation. Chematica network consists of nodes and edges (like Google Knowledge Graph)  which represent molecules and reactions respectively. Chemical rules have been built into this network, enabling it to predict optimal routes to synthesize a compound of interest  (Ref.2,3,4). The three papers in Angewandte Chemie describe three use cases of Chematica : 1. One pot synthesis for multi-step syntheses (Ref. 5) 2. Optimizing existing syntheses (Ref. 6) 3. Identify synthetic routes leading to harmful substances  (Ref. 7) These papers demonstrate how application of Chematica knowledgebase could be useful in bringing down synthesis costs in pharma and chemical industry.  This database was the fruit of a decade of research work and has the potential to be a path breaking technology for chemistry.  A comparative study of current databases of chemical reactions and Chematica would enable chemists to get the best out of them. Read the full article.    

Keywords:  Chematica • Knowledge Graph

---

1 Trackback

• Künstliche Intelligenz - Stellen für Chemiker