During my sophomore year of college, I built a portable device to analyze the diffraction patterns of C. elegans worms for Dr. Magnes at Vassar College.
The Context.
Date: fall 2014 - spring 2015
Location: Vassar College, Poughkeepsie, NY
One-year of biophysics research under the guidance of Professor Jenny Magnes.
My Role.
Arduino coding
Circuit design
Prototyping
Data analysis
The Science
Dr. Magnes’ research focuses on using the properties of light to understand how tiny organisms move without actually have to see them.
We did this by shining a laser light through a cuvette containing a worm floating in water, which creates diffraction patterns (right).
Worms typically thrashed in one of two ways: “wildtype” or “rolling” (bottom right).
Sample of worm motion (left) versus associated diffraction patterns (middle and right)
Experimental set-up
The Data
I performed a time series analysis to determine if the nematode’s motion follows a pattern or demonstrates chaotic behavior. To this end, I used three computational techniques in Matlab:
Lag Plots.
Lag plots display the value of the data at time (t) versus the data at time (t – lag), where lag is a fixed time displacement. This method allows researchers to infer dynamical attractors from observations.
In other words, if the system tends to move towards specific values, we will be able to see it in the lag plot. If there is no pattern (e.g. as in chaotic motion), no pattern will be observed.
Density Plots.
Density plots give information about the number of times point (x,y) appears in each plot using color. By analyzing how much of the plot contains nonzero vs. zero values, the area the worm uses to move around in can be statistically described.
Poincaré Plots.
Poincaré plots graph each value (point at time t) against the next chosen value (point at time t + lag). They can be used to describe how a system evolves over time and visualize data variability.
They have two basic statistical descriptors, SD1 and SD2, that measure dispersion perpendicular and parallel to a 45-degree diagonal reference line called the “line of identity.”
The results of our experiment showed non-chaotic behavior and provided insights into locomotive differences between roller and wildtype worms. They indicated that this method could be used to identify the worm-type and even predict future worm motion.
Taking it to-go
My main research goal was to build a portable device that could take data in the field, allowing researchers to study microorganisms in their natural environment. It would have to collect data, cut down on noisy samples, and save the data for later analysis.
The first challenge in this was filtering the data, for which I needed to build a circuit that would remove erroneous frequencies. Although, at the time, I’d never worked with Arduinos or circuits, I was eager to dive in and learn. I used online resources and a coding book to pick up the basics, then built a bandpass filter comprised of high and low pass filters in series. This configuration was designed to filter out noise but keep the sample frequencies we needed.
In the video below, I demonstrate signals outside of the range being filtered out once the bandpass filter is turned on.
Prototyping
The Schematic.
Once the bandpass filter was working, I incorporated each additional component in step-by-step. The final prototype consisted of a battery-powered Arduino Uno connected to a photosensor in turn attached to the bandpass filter. Frequencies were displayed on a small oscilloscope screen and data were saved to an SD card.
The Prototype.
In the end, the prototype was able to take in light sensor information, filter out extraneous frequencies, plot data in real-time on an oscilloscope, and save to an SD card for later analysis in the lab.
Next Steps.
I presented the prototype at the senior’s research symposium. For the work, I was selected as one of three Undergraduate Research Summer Institute (URSI) fellows to continue the research. Although I decided to research solar cell materials at Colorado School of Mines that summer instead, I always looked back on this project fondly, as it was my first real taste of engineering design.
A good friend of mine and another URSI fellow, Christine Silveira, continued the project, finishing the prototype (shown above) and successfully using it to analyze C. elegans diffraction patterns!
Why it Matters
The methods developed in Dr. Magnes’ lab allow researchers to study in complex, constrained and unconstrained 3D environments. Conversely, traditional microscopic methods require that the subject be confined to the image plane. This technique can be used to elucidate and model the neural networks that drive C. elegans motions and can also be used to study the impact of genetic defects on locomotion.
On a personal note, I consider this project as one that influenced my path towards design engineering. Building prototypes, testing out ideas by trial and error, learning to code, and seeing it all come together was such a fun process that I couldn’t help but seek out more as I moved forward.