Sandhya Govindarajan

Computational Biologist 1 @Ohsu Knight Cancer Institute

Portland, OR, US

Customized and tuned a Deep Learning model (VQGAN) to analyze 10B+ images with 40 biomarkers; achieved a Spearman correlation of 0.9, enabling a 60% reduction in the marker panel size, reducing costs associated with multiplexed staining.Utilized AWS to download 2 TB of data containing 5 billion+ 3D confocal images. Customized a Variational Autoencoder (VAE) to detect circulating hybrid cells (CHCs) that can be used for early cancer prognosis.