about?
I’m a final year undergrad studying Computer Science and Engineering at VIT Chennai. I’m also studying Data Science at IIT Madras. My areas of interest span machine learning, genomics, and disease diagnostics. Interested in learning more? Read about my projects →
Currently, I lead engineering at Cogny, where we’re building next-gen diagnostic platforms using immersive technologies. Our work aims to revolutionise how we approach cognitive and behavioral health assessments.
I have research on early blindness detection using deep learning, modular components for assistive technology systems, and designing a multimodal framework for cancer mutation detection through histopathological analysis. Check out my research →
When I’m not caught up in code or research, I like capturing moments — there’s something special about freezing a frame in time; pictures of a poem if you will. A quiet rebellion against time, so to say.
looking into
- The potential of transformer architectures in predicting protein-protein interactions, particularly how self-attention mechanisms can capture complex molecular binding patterns.
- Emerging applications of graph neural networks in modeling cellular regulatory networks and metabolic pathways.
- Novel approaches in deep learning for drug-target interaction prediction, particularly attention-based models that can handle both 3D protein structures and molecular sequences.
intrigued by
Diving into computational genomics, I’m currently learning the concepts of spatial transcriptomics - a technology that preserves spatial information while measuring gene expression. While traditional RNA sequencing gives us expression data, spatial transcriptomics shows us where in the tissue each gene is active. This spatial context is crucial for understanding how cells interact and organize themselves in tissues. I’m exploring how deep learning can help make sense of this data, especially using graph neural networks.
The most exciting part is seeing how we can combine spatial transcriptomics with other data types like protein measurements. I’m learning to work with transformer architectures that can handle this multi-modal data, though the computational challenges are significant. My newest project involves building a pipeline that can process spatial transcriptomics data while preserving its biological interpretability. There’s still much left to learn, but it’s thrilling to be able to work at the intersection of deep learning and genomics.