There are roughly 1084 atoms in the universe. This number is significant, because it places a bound on the amount of information that can be stored in matter, or the number of experiments that can be performed at any given time. This is important for biologists who want to understand how systems work by enumerating sequences, which can grow larger than this number very quickly.
The approach to "understanding biology" that we are pursuing acknowledges that enumeration, albeit useful, is not a feasible in the long run. Our approach is to first make many measurements of developing biological systems using sequencing. Model these data using deep learning, and generate new instances that have been previously unexplored. Our work is focused primarily on developing sequencing based measurements, due to the scale that can be achieved.
The spatial organization of cells is apparent in any multi-cellular system. Some of this structure is emergent due to the physics, while other aspects are encoded in the genome creating the sterotyped and choreographed process of development.
We are building new techniques to make rapid and large spatial measurements, with the goal of documenting and modeling how developing embryos form. Our ultimate goal is to use this information to imagine, generate, and build new tissue architectures.
A new area in the lab. We're working towards developing sequencing-based assays to measure and monitor biochemical reactions. We are looking to take advantage of the scale and throughput of sequencing experiments, to understand how variation in sequence (RNA or protein) leads to changes in biophysical or biochemical parameters of biological complexes.
Single-cell sequencing has had two profound impacts in biology. First, it has furnished a rich and unified language for the compilation and comparison of cells across various tissues, organisms, and species. Second, it has increased experimental throughput. By using each cell as a container, large genetic screens can be conducted followed by high-content, genomic readouts as the phenotype.
For both of these applications, we are developing new single cell co-assays. Co-assays are designed to measure multiple genomic observations within the same cell. This simultaneous measurement provides an explicit connection between the observed modalities and allows for more robust causal inference.
The Srivatsan Lab is fascinated by embryonic development. Over a short time, cells divide, move, and take form guided by an internal program. For those that have never witnessed this process, we recommend watching Becoming by Van Ijken, which documents the development of a salamander. In this compact video, invisible and diverse cellular and biochemical processes are rapidly occurring to reshape a rough mass of cells into an organism. A few questions we're interested in include:
1. When are cellular fates plastic during development?
2. How many progenitors give rise to each cell type?
3. What dies with cells that constitute lineal dead ends?
4. What would a new cell type look like and can we make it?