BioSB Young Investigator Award winner 2022

The selection committee of the BioSB Young Investigator Award 2022 unanimously chose for Janani Durairaj as the winner of the BioSB Young Investigator Award 2022! 

Janani will receive the award during the award ceremony on 28 June 2022, 16:30 and she will give an honorary lecture. 

Janani Durairaj

Torsten Schwede Computational Structural Biology Group
University of Basel, Switzerland

Biography

Janani Durairaj is a computational biologist working at the interface of protein structural bioinformatics and machine learning. She completed a dual major, a Bachelor of Engineering in computer science and a Master's in biological science, from BITS Pilani, India. During the last year of her masters, she traveled to Wageningen University for an internship in protein-protein interface prediction at the Bioinformatics group. In August 2017, she then started her PhD in this group supervised by Dr. Aalt-Jan van Dijk and Prof. Dick de Ridder. Her PhD research involved development of novel approaches combining homology modeling, structural and chemical feature extraction, and machine learning. This enabled her to predict and understand which fragrance molecules are produced by different terpene synthases, which in turn informed experimental studies by collaborators in the lab. Along the way, she developed computational protein structure representations drawing inspiration from the fields of robotics and computer vision. She defended her PhD thesis "Computational approaches to discover novel enzymes for fragrance and flavour" in September 2021 and obtained her degree at Wageningen University with the distinction cum laude. She is currently a postdoctoral researcher with Prof. Torsten Schwede at the University of Basel, working on protein representations and benchmarking for computational drug discovery.

Presentation: Structure is the new sequence

The relationship between protein sequence and function is highly intertwined with the intermediate protein structure. Dramatically improved protein structure prediction and the ever-growing number of experimentally solved protein structures have led to an age where large-scale protein function exploration studies can now incorporate a gold mine of structural information instead of or along with the typically used sequence information. We develop novel computational representations based on predicted and experimental protein structures which consider various perspectives such as the topological shapes of three-dimensional substructures, the inherent dynamic nature of proteins, their many interactions with ligands, the underlying physicochemical and electrostatic properties of the amino acids involved, and so on. These representations are used in machine learning algorithms to link proteins to their function, while pinpointing residues and structural regions most likely to be responsible for carrying out this function. We apply these approaches across different protein families, and specifically to the family of terpene synthases, which share a common structural fold across wide sequence diversity and are responsible for the large variety of plant fragrances around us.