Research Interests: Cells in the human body can be subdivided into transcriptomically distinct types, which correspond well to classical distinctions based on morphology, physiology and connectivity. Knowing the full compendium of cell types in a given tissue, and having reliable genetic markers for each type will enable cellular circuits to be studied at unprecedented molecular resolution in health, development, disease and across different species. Together with experimental collaborators, I am applying large-scale single-cell transcriptomic analysis of the vertebrate retina as a model system to address these issues. Building on initial “proof-of-principle†studies (Macosko et al.,
Cell, 2015 ; Shekhar et al.,
Cell, 2016; Pandey, Shekhar et al.,
Current Biology,2016), I am poised to complete a comprehensive neuronal atlas of the mouse retina, which will be among the first of its kind for any brain region. Using state-of-the-art machine learning techniques, we have defined molecular signatures for a total of >120 neuronal types, including 3 photoreceptor, 1 horizontal, 15 bipolar, 45 ganglion, >60 amacrine and ~10 non-neuronal cell types, confirming a large number of them histologically. Using the mouse “retinome†as a foundation, I am to pursue two directions.
First we are performing large-scale single-cell surveys of two other vertebrate retinas that differ from that of the mouse in key aspects: (a) macaque retina, which contains the fovea, a structure responsible for high acuity vision in primates that is absent in mice, and (b) zebrafish retina, which unlike mice and primates, can regenerate following injury. Using molecular signatures, I aim to computationally compare the taxonomy of retinal neurons across species, and identify “orthologous†neuronal types. Beyond uncovering fundamental biological principles, this also has implications in revealing molecular pathways associated with regeneration, and the unique properties of vision in primates.
Second, we are transcriptiomically profiling neuronal and non-neuronal cells of the mouse retina following physical injury to the optic nerve. By comparing changes in gene expression among types following injury, we can identify early transcriptional signatures that correlate with, and may underlie, selective resilience. In addition, we are identifying injury responses of interneurons and non-neuronal cells (glia and immune cells), all of which have been shown to influence survival. This project has the potential to uncover strategies to reverse neuronal death in the context of injuries and neurodegenerative diseases like glaucoma, a leading cause of blindness in the developing world.
Teaching Interests: I believe that a sustained teaching practice is a fundamental cornerstone of scientific growth. My doctoral and postdoctoral experience has positioned me to teach a diversity of courses. I describe key highlights of my teaching experience: First, as a graduate teaching assistant for the graduate course on thermodynamics and statistical mechanics at MIT Chemical engineering (Instructors : Arup Chakraborty and Bradley Olsen), I was awarded the best TA award in 2011. Second, I have been a guest lecturer in the MIT Biological Engineering course "Fundamentals of Biological Networks" targeted towards advanced undergraduates and graduates. Third, as a post-doctoral associate, I have designed and delivered short course introduction to single-cell genomics. For e.g. I was invited to deliver a short course on single-cell transcriptomics at the Evomics 2018 conference in Cesky Krumlov, Czech Republic. I believe that I have a strong foundation to teach undergraduate and graduate courses on thermodynamics, statistical mechanics and reaction engineering. I would also like to design graduate electives that brings together physical biology and statistical learning techniques drawing from my own research in single-cell genomics.
