Abhay Kshirsagar

Mumbai · Maharashtra · India · (+91) 9321201230 · ask19ms172@iiserkol.ac.in

I'm 5th year Integrated MS Student Majoring in Biological Sciences and Minoring in Computational and Data Sciences at IISER Kolkata. Currently I am completing my final year Master Thesis project at IIIT Hyderabad under the supervision of Prof. Deva Priyakumar. I'm interested in studying Biological Phenomenas using Machine Learning and Computational techniques. You can find my CV here .

Publications

Conformational Ensemble of the NSP1 CTD in SARS-Cov2: Perspectives from the Free Energy Landscape (Accepted In pre-proof)

Dutta, P., Kshirsagar, A., Bibekar, P., Sengupta N,. 2023.
DOI: 10.1016/j.bpj.2023.02.010
In this Study we investigate the conformational ensemble of NSP1-CTD of SARS-CoV-2 virus by leveraging Enhanced Sampling Method like Replica Exchange Molecular Dynamics and gain insights on its conformational landscape by using methods such as Expectation Maximized Molecular Dynamics

Experience

Master's Thesis Project

IIIT Hyderabad
  1. Currently I'm building Machine Learning models that can Predict Amyloidogenic Hotspot regions in Protein Sequences.
  2. Learned about Large Language Models for Protein Sequence Encoding.
  3. Exploring how Physical and Chemical properties can be incorporated in LLMs to improve the prediction accuracy.
June 2023 -- Present

Research Internship

IISER-Kolkata
  1. Learned about Molecular Dynamics Simulations for Biological Systems
  2. Learned and worked on some Deep Learning Frameworks for studying conformational landscape of protein like Time Lagged AutoEncoders (Noe. at. al. ) and AutoEncoders (Mondal et. al. )
  3. Worked on NSP1 protein of SARS-CoV-2, specifically its CTD which behaves like an IDR. We studied its conformational ensemble using an enhanced sampling method REMD and MD. We employed Expectation Maximized Molecular Dynamics to gain insights on its conformational landscape.
November 2021 -- December 2022

NLP Researcher

Wavel.ai
  1. Worked on the core Data Science and NLP Engineering Team at wavel.ai.
  2. Learned how to create train and evaluate models and create pipelines for Text-To-Speech
  3. Learned about various state of the art TTS models for multilingual TTS and Voice Cloning Some of the models I Implemented were based on following architectures
June 2022 -- August 2022

Summer Research Project Intern

IISER-Kolkata

Completed a two month research Internship under Dr.Jayasri Das Sarma (Dept. Biological Sciences) where I studied the effects of SARS-CoV-2 on humans with Murine-CoV as a model. Also gained insights on the lab methods for studying animal models for human diseases.

June 2020 - August 2020

Projects

Solar Power Prediction

SHELL.ai hackathon
My Submission to the Shell AI Hackathon for predicting Global Solar Irradiance and Cloud Coverage. Implemented Long Short Term Memory Networks on previous Data to Forecast Target Variables. My Team were selected to be finalist among Top 30 Teams. The Approach and the Certificate is provided in the Repository Below

 Project Link  Certificate

Wind Farm Layout Optimization program

SHELL.ai hackathon
My Submission to the Shell AI Hackathon for finding Optimum Wind Turbine Layout for maximum Annual Energy Production. In the Project I used a Stochastic Algorithm called Particle Swarm Optimization Algorithm (PSO)

 Project Link

Flappy Bird AI

Personal
Made an self Supervised AI that Plays the famous game Flappy Bird. The Program is build using NEAT Algorithm provided by the neat library.

 Project Link

WELEARN BOT

Personal Team Project
A cli application built using Python for interacting with our Institutes Course Content Website on Welearn. The program uses Moodle Web Service API for downloading course content and making Google Calendar reminders for Assignment or Test Submissions.

 Project Link

Education

Indian Institute of Science Education and Research Kolkata

Intergrated BS MS
Majoring Biological Sciences and Minor in Computational Data Sciences

GPA: 8.59

August 2019 ~ Present

Skills

Programming Languages & Tools
FrameWorks
  • Android Development - Flutter and Dart
  • Web Development - Node and React (Limited Proficiency), LAMMP Stack
  • ML FrameWork - Tensorflow, PyTorch, PyTorch Geometric (Limited Proficiency)
  • Python libraries - Numpy, ScikitLearn, Pandas
  • Molecular Dynamics - GROMACS, VMD
  • Scientific Computing - MATLAB, Simulinks, OCTAVE (Limited Proficiency)
  • Database - SQL, NoSQL (Limited PRo)

Awards

National HPC Hackathon

Hackathon
Part of the winning team in the National HPC Hackathon organised by Amazon Web Services and Intel in association with the Office of Principal Scientific Advisor to the Government of India.
December 2021

INSPIRE scholar 2019

Scholarship
Scholarship Program for Students pursuing Basic Sciences and showing Academic Proficiency.

Kishore Vaigyanik Protsahan Yojana (KVPY) SB scholar

Scholarship
Scholarship Program for Students pursuing Basic Sciences and showing Academic Proficiency.

RANK: 56

December 2019 ~ present

Shell AI Hackathon

Hackathon
Me and my Team were recognised as the finalist (ie in Top 30 Teams) in the Shell AI Hackathon for Sustainable and Affordable Energy. The Link to Project and Certificate are provided below.

Certificate , Project

November ~ December 2021

Certifications

Neural Networks and Deep Learning

Coursera
Course taught by Andrew Ng from DeepLearning.ai. The Course gives an introduction to Basic Neural Networks and goes over Regression and Classification Examples using Python and its libraries like Numpy.

Certificate

Machine Learning

Coursera
Course taught by Andrew Ng from Stanford University. The Course goes over various Supervised (Regression, SVM, Kernels, NN) and Unsupervised Learning Algorithms (Clustering, Dimensionality Reduction, etc). It also dwells deep into best Practices and builds these foundations using plenty examples with MATLAB / OCTAVE.

Certificate

Improving Deep Neural Networks: Hyperparameter Tuning, Regularization and Optimization

Coursera
Course taught by Andrew Ng from DeepLearning.ai. The Course goes over standard Neural Network techniques and dives into how to improve the performance of the Deep Neural Network Blackbox. The Course uses many Examples to understand the improvements each modifications bring using ML FrameWorks such as Tensorflow.

Certificate

Convolutional Neural Networks

Coursera
Course taught by Andrew Ng from DeepLearning.ai. The Course goes over CNN and how it revolutionized the field of Computer Vision. The course shows how to build CNN and Residual Networks to build projects such as Style-Transfer, Image and Video Detection, Image Semantic Segmentation using ML Framworks like Tensorflow.

Certificate