This summer, I investigated sleep spindles in the Weber Lab, which does sleep research.  Sleep spindles are events that last for around half a second to two seconds, and have many interesting properties.  Spindles are associated with learning and memory, and occur in a type of sleep called NREM (non-REM) sleep.  My project looked at sleep spindles from many different angles to help our understanding about sleep.

The first step of my research was to point out where sleep spindles occurred in sleep recordings, to prepare the data for analysis; this process was called annotating.  The sleep recordings were often tens of thousands of seconds long and took hours to annotate, and after several annotations that I did which took me many hours over a couple weeks, I decided to build an algorithm that would do the annotation for me, which would be faster.  And instead of taking several hours for annotating a single recording, this sleep spindle detection algorithm took a couple minutes!

With my sleep spindle detection algorithm, I was now ready to analyze the data, and patterns began emerging.  I built a graphing module which I named grumpy after the Python libraries and modules which I used extensively called numpy, pandas, scipy, and sleepy. I called my graphing module grumpy to reflect my mood while doing a sleep spindle annotation before the spindle detection algorithm was built. 

Looking at several data sets, including fear learning, social defeat, chemogenetics, and optogenetics, I compared sleep spindles between different experimental conditions.  I also looked at how sleep spindles fit into sleep, without any experimental condition; this was called sleep architecture.  In sleep architecture, I investigated the relationship of REM and spindles, NREM and spindles, and wake and spindles, and saw many interesting patterns!

From this research, I learned a lot of programming, primarily python.  As the summer progressed, I began to feel more and more comfortable with python.  Although the coding process was sometimes arduous, especially when debugging, the resulting graphs were beautiful and I couldn’t wait to stumble upon the next revelation.  Each graph led to another area of exploration.  I also learned a lot of sleep science by reading articles from the literature, but the primary focus of the research was data driven. 

At the beginning of the summer, Dr. Weber, my mentor, when showing me sleep spindles in the EEG for the first time, described them as “beautiful.”  And now, I have come to appreciate them as well.