Vanderbilt holds an annual Three Minute Thesis competition (original from University of Queensland where graduate students from all departments get the chance to describe their research to a lay audience. The goal is to be able to communicate difficult concepts in simple terms and in an efficient amount of time. I decided to participate again this year so I figured I share my draft speech to give you an overview of my project. At the bottom, I have my version from three years ago as well for comparison. As always, I’m happy to hear feedback!

2017 Version: If The Glove Fits

Imagine it’s a snowy winter day in the middle of March. You’re going for a walk but first you need to put on some gloves. You probably don’t think about the process of putting on the gloves, it is muscle memory. There might have been a time when you had to learn it as a child, but it was likely very intuitive. Now consider it from the perspective of a computer.

The computer has no concept of human body parts, but rather it relies on bending flexible joints to the shape it wants. There are two joints in each of your fingers in addition to the base that allows it to rotate. Your wrist is also flexible giving a total of 16 moveable joints in one hand. The computer has to determine how to manipulate the hand to the shape for putting on the glove bottom, and then changing it to finish the process. If you factor in the shape and size of the glove as well, it becomes increasing complex for a computer to determine how to put on the glove. Discovering new drugs faces the same challenge.

In order to find a new drug, we have to find a molecule that will bind a disease target. A cancer drug for example has to bind a protein essential to the growth of cancer cells. Think of the drug as your hand the disease target as the glove. Just as we have to do lot shape changing and hand movements to put on the glove, we have to bend molecules of all shapes and sizes to fit into a drug target. An added complication is that in medicine, you do not always know what the target looks like!

My research is centered on writing computer programs to accomplish the task of fitting molecules, or as we call it, docking. My labmates and I start by finding ways to simplifying drug molecules and disease targets to the essential parts. This makes it much easier for the computer to handle. You can still put on a glove if you kept all your fingers straight! We also develop ways to move our virtual molecules up, down, left, right and rotating in different directions.

Lastly, it isn’t useful to know how to do a task if you don’t know when you have completed it. We have to teach the computer to recognize when the molecule is properly docked into the drug target. To do this, we look for similarities in all known ways that proteins and molecules interact with each other and train the computer to recognize these commonalities. Once we have a computer model of the molecule docked to the target, we can then use it to design better drug molecules!

2014 Version: Dock It Like It’s Hot

I’m a big Doctor Who fan so let’s time travel for a moment back to Elizabethan England, to the height of Shakespeare, the Renaissance, and crazy medical practices. If you had a sore throat, you might be prescribed slime from a garden snail. A toothache is treated with paste made from a mouse. Yuck! It’s no wonder these remedies didn’t work because doctors had little knowledge of how medicines worked. Fortunately, we’ve since learned a thing or two about both WHAT we should take and about HOW those drugs worked. Modern drug discovery though still has its own challenges, thanks to a few very large numbers.

It is estimated that there are 10^33 potential drugs out there. That’s 1 followed by 33 zeros! If we were to make a single pill for each of those drugs, the total would weigh as much as the sun.

Understanding how a particular drug works isn’t easy either. We know drugs act by binding to specific microscopic machinery in our body called proteins, similar to the way a key opens a particular lock. What makes it difficult is that there are 100,000 different types of proteins in our body. Imagine trying to figure out which lock a particular key opened in a building with 100,000 doors. What a nightmare. This makes it impossible for scientists to test everything by hand.

Enter the computer. The smart phone in your pocket is a computer capable of performing a billion calculations every second. There are supercomputers out there that are even faster. Why wouldn’t we harness this immense resource to help us find and understand the next generation of therapeutics?

Researchers have developed two separate set of computer software for drug discovery. Virtual screening help us filter down the near infinite list of possible drugs to a testable list of thousands. Molecular docking helps us understand how a specific drug unlocks a target protein.

My research focuses on bridging the gap between the two sides. I want to integrate our knowledge of WHAT drugs work from screening to help us figure out HOW they work through docking. For example, this morning I woke up with allergies and had to take a Claritin before coming here. Some of you might have had a similar experience but took an Allegra or a Benadryl instead. If we compare and contrast these drugs, we can gain insight into on how they work to treat the same symptoms. Let’s tackle the challenges of modern drug discovery by combining the WHAT question with the HOW question.

I like to leave you with some lyrics from the Snoop Dogg song that inspired the title of my presentation, “You should think about it, take a second. Matter fact, you should take four.  “