Skip to main content
Contact Info
Thenmalarchelvi Rathinavelan
Department of Biotechnology, Indian Institute of Technology

Abstract OpenTox Asia 2019 

A pattern recognition approach to predict a protein’s like or dislike when it meets a metal

L Ponoop Prasad Patro*, C Sathyaseelan*, Kripi Tomar* and Thenmalarchelvi Rathinavelan

Metal ions play a crucial role in stabilizing protein structures as well as governing their complex biological functions. Metal ion binding proteins, namely, metalloproteins play a central role in signal transduction, transport, storage and catalysis. Metalloenzymes, a subclass of metalloprotein catalyzes complex chemical transformations that are essential for cellular functions. Nearly one-third of proteins bind with metals either to facilitate structural stability or to regulate the aforementioned biological functions. The functional diversity of proteins is mainly driven by the type of metal ion that coordinates with the protein. Metal ion...protein interaction also has advantages in nanotechnology and therapeutics. Despite the realization of the importance of metal ion and protein interaction in biological reactions, a very few studies are available regarding the sequence and structural conservation of proteins and the concomitant choice for the metal. This mandates a systematic investigation to characterize the metal ion coordination with proteins. We have carried out a rigorous statistical analysis on 31,220 protein structures from protein databank (PDB) that are coordinated with 52 different metals and derived “metal ion ... amino acid” interaction matrices. The matrices eventually reflect the binding preference of 20 amino acids to 52 metals. Similarly, a thorough analysis has been made on 37 metal binding protein classes to derive the amino acid sequence patterns and their preference for metals. This has been implemented as an automated web-server that can predict the metal binding region in proteins using both structure and sequence. Thus, the statistical and pattern recognition algorithms implemented here would significantly help towards accurate prediction of uncharacterized proteins, identification of metal binding regions in NMR and X-ray derived structures, metal ion mediated drug binding as well as tailor-making of artificial metalloproteins for nanotechnology application by utilizing the beauty of nature's selectivity.