D.A.R.L.I.N.G

Deep leARning for chemicaL Information processinG

Vast quantities of scientific information are hidden in primary scientific publications and not available as curated data in scientific databases. Making such information publicly available to support open science and open innovation is a challenge that has to be solved. In this dissertation, state-of-the-art deep learning models for optical chemical structure recognition and chemical information processing have been implemented to rediscover this information and retrieve it automatically.

STOUT (Smiles TO iUpac Translator)

STOUT: SMILES TO IUPAC Translator is built using the same concept as a Neural Machine Translation(NMT). STOUT is initially trained on a subset downloaded from PubChem containing 30 Million SMILES and 60 Million SMILES. which got converted into SELFIES using the SELFIES package. The same set of SMILES also was converted into IUPAC names using ChemAxon "molconvert", a command-line program in Marvin Suite 20.15 from ChemAxon (https://www.chemaxon.com). Later the textual data was converted into TFRecords(Binary files) for training on Tensor Processing Units(TPUs).

MSC (Molecule Set Comparator)

Molecule Set Comparator (MSC) is designed as an application that enables a user to do a versatile and fast comparison of large molecule sets with a unique inter-set, molecule-to-molecule comparison, for the original set and a predicted set of molecules obtained by machine learning approaches. The molecule-to-molecule comparison is based on chemical descriptors, which are included in the Chemistry Development Kit (CDK), such as Tanimoto similarities, atom/bond/ring counts, and physicochemical properties like logP. The results are presented graphically and summarized by interactive histograms that can be exported in publication quality

COCONUT COlleCtion of Open NatUral producTs

I was part of the data curation work and desinged the logo for the project

A rule-based method for systematic conformer generation

A Java based to tool to explore and generate the conformers of a bioactive molecule systematically.

Friedrich-Schiller-Universität, Jena, Germany.

In 2021 I completed and defended my Ph.D with a "Summa cum laude" grade.

Savithribai Phule Pune University, Pune, India.

I completed my Master’s degree in Bioinformatics, in Pune, India. For my Masters’ thesis, I worked primarily in cheminformatics.

Bangalore University, Bangalore, India

I completed my Bachelor of Science in Biotechnology, in Bangalore, India.