Project: Cell-to-Cell Communication for Distributed Computation in Bacteria.
I want to build a computer using bacteria.
Any function can be expressed with logic gates (like AND, OR, NOT, etc.), and we can implement logic gates with genetic circuits. So in theory, we could implement any function in bacteria.
Unfortunately, the state-of-the-art is that we can only fit a handful of logic gates into each cell. To implement more complex logic circuits, we need to distribute the gates across multiple cells, and we need a way to send signals from one cell to another.
My final project will focus on implementing and characterizing cell-to-cell communication.
I was primarily inspired by the work of Jai Padmakumar, who for his thesis project at the Voigt Lab, implemented a reduced version of the MD5 hash function in E. coli (!)
I’ve long been interested in cryptography, so it was very exciting to see someone combining cryptographic functions with bacteria.
When I first heard about genetic circuits, I had imagined that we could implement arbitrarily complex computation in cells. To a software engineer, even the simplest function would take thousands of millions of gates to express. Unfortunately, it turns out that you can only put maybe 3 or 4 gates into one cell.
(In the Week 7 lecture, Ron Weiss said that functions on 5 inputs or so is sufficient for any biological engineering application. That may be adequate for curing cancer, but it’s deeply unsatisfying to computer programmers used to much more complex functions.)
To implement more complex functions, you need to partition the gates across multiple cells, and we need a way for the cells to talk to each other. In an electrical circuit, the signals between gates are carried as voltages on wires. In cells, the signals are small molecules that cells secrete and other cells can detect.
For my final project, I want to try implementing cell-to-cell communication in E. coli.
Here is another paper where they distributed a logic circuit across multiple cells. Whereas Padmakumar cultured his bacteria strains separately and transferred the signalling molecules from one strain of bacteria to another by hand, in this paper, they co-cultured the different strains of bacteria to get a single result at the end:
For the implementation of the communication devices, this paper has a lot of optimized genes and pathways for signalling molecules: