Priyanshu Kumar
Computational Structural Biology | Network Science | Stochastic Modeling
Second-year Biotechnology undergraduate at Chandigarh University developing reproducible computational frameworks for protein structure evaluation and network analysis. Research focus on integrating machine learning with physics-based modeling for structural biology applications.
Discover My Journey
About Me
Bridging computational biology, network science, and stochastic modeling to decode biological systems
Research Background
I am a second-year Biotechnology undergraduate at Chandigarh University with research interests in computational structural biology and network science. My work focuses on developing reproducible computational pipelines for protein structure evaluation, investigating the application of machine learning methods to structural biology problems, and exploring network-theoretic approaches to biological systems.
Through independent research, I have reproduced and extended the AF2Rank methodology for AlphaFold-based structure evaluation, analyzing over 145,000 Rosetta decoy structures. This work demonstrates how machine learning models can quantitatively assess protein structure quality and explores connections between model confidence patterns and network-theoretic properties of protein systems.
Research Interests
My research interests lie in developing integrative computational frameworks that combine deep learning, network theory, and physics-based modeling for structural biology applications. I am particularly interested in improving protein structure prediction methodologies, understanding network properties in biological systems, and applying stochastic modeling approaches to computational neuroscience and systems biology.
CGPA 8.13/10
First Year Academic Performance
First Author
bioRxiv Publication (Under Review - Springer)
Tech Invent 2024
Second Place - IoT Water Quality System
Education
Bachelor of Engineering
Biotechnology • Chandigarh University
CGPA: 8.13/10 (First Year)
Focus: Computational Biology, Network Science
Duration: Aug 2024 – Jul 2028
Technical Skills
Programming & Scientific Computing
Machine Learning & Modeling
Computational Biology
Data Analysis & Network Science
Research & Publications
Advancing computational biology through reproducible research and innovative methodologies
Research Focus Areas
Protein Structure Prediction & Evaluation
1 publication
Computational methods for evaluating protein structure predictions using AlphaFold and Rosetta. Development of quantitative confidence metrics and reproducible assessment frameworks for structure quality evaluation.
Network Science & Stochastic Modeling
1 publication
Investigation of network-theoretic properties in biological systems, including small-world network characteristics and stochastic transition patterns. Applications to protein structure networks and their relation to model confidence metrics.
Computational & Systems Biology
1 publication
Development of reproducible computational pipelines for structural biology research. Focus on creating scalable workflows that integrate machine learning approaches with traditional physics-based modeling methods.
Research Directions
My research aims to develop integrative computational approaches that combine machine learning, network theory, and physics-based modeling for structural biology applications. Building on current work in AlphaFold model evaluation and network analysis of protein systems, I seek to contribute to advancing structure prediction methodologies and exploring their applications in computational neuroscience and systems biology.
— Focus on reproducible, scalable computational frameworks for biological research
Publications & Research Output
AF2Rank Revisited: Reproducing AlphaFold-Based Structure Evaluation and a Hypothesis for Context-Aware Refinement (CAR-AF)
Author: Priyanshu Kumar
Journal: bioRxiv (Submitted to The Journal of Supercomputing - Springer)
Technical Projects
Research and development projects in computational biology, machine learning, and software engineering
AF2Rank Reproduction & Extension
Reproduction and extension of the AF2Rank methodology for evaluating AlphaFold structure predictions, analyzing 145,000+ decoy structures across 133 protein targets.
Physics-Informed Neural Networks for Protein Dynamics
Investigation of hybrid neural network architectures that incorporate physics-based constraints for protein modeling, in collaboration with mathematics faculty.
Research Collaboration
Open to discussing research collaborations, project opportunities, and academic exchanges in computational structural biology and related fields.
Contact
Open to discussing research opportunities, collaborations, and academic exchanges in computational structural biology
Get In Touch
For research collaborations, questions about my work, or discussions regarding opportunities in computational structural biology, please feel free to reach out.
Quick Actions
Send a Message
Please use the form below to reach out. I typically respond within 24-48 hours.
Response Time
I typically respond within 24-48 hours during business days.