Season 3, Ep. 2

Thy Pham: There’s this concept of being a very talented scientist, you know, I thought it was that you had to be super sharp and super smart. But nowadays I also really appreciate patience and persistence.

[“Something Elated” begins]

Conor Gearin: Welcome to BioGenesis, where we get to know a biologist, where they came from, and where they’re going next. I’m Conor Gearin from Whitehead Institute —

Raleigh McElvery: And I’m Raleigh McElvery from the MIT Department of Biology.

Gearin: This season, we’ll introduce you to three graduate students who are “Breaking the Mold.”

[“Something Elated” fades out]

McElvery: Over the course of their research training, these biologists have shattered the pre-conceived notions they grew up with about what scientists look like and where scientists come from.

Gearin: Today, you’ll meet Thy Pham.

McElvery: After moving to the U.S. from Vietnam at age 20, she learned English while delving into biophysics and computer science courses, opening a door to a career that she hadn’t realized was available to her.

Gearin: At Whitehead Institute, she’s applying artificial intelligence to predict how cells fine tune gene expression using short sequences of RNA called microRNAs.

[“Brek PKL” begins]

Pham: My name is Thy Pham, I’m a fifth-year graduate student in the Bartel lab. So I grew up in a small fishing town by the beach in Vietnam. So it’s really, really beautiful, it’s touristy. My parents had at first a small business, and then it kind of grew into like a medium-sized business. So my mom helped my dad with the business, and they just buy like sea shells and make them into souvenirs and export them to other parts of the country or other countries.

[“Brek PKL” fades out]

So I read a lot of comic books growing up. Manga was really big in Vietnam back then, but I also read a lot of just fiction. I tend to read like whatever is around me and my parents have a lot of Buddhism texts, so I remember reading them a lot as a kid.

Gearin: Thy was also drawn to math, because of its elegant simplicity.

Pham: I just really liked math when I was growing up. Like, every Saturday, I would wake up very early and do all my math homework, which is not a normal thing for a lot of kids. A lot of the time things don’t make sense, like human behaviors in real life don’t sense at all, in my opinion as a kid. So I think for me that sense of very pure and abstract thinking was really beautiful and comforting.

In sixth grade, I got a book about like the history of math that focused on mathematicians. So I think that was my first exposure to what is a scientist, what is a mathematician. And I actually really appreciate how diverse those stories are.

[“Dowdy” begins]

But I think that one of my most formative moment thinking about scientists is when I learned about Mendel and genetics.

McElvery: Gregor Mendel, a 19th century Austrian monk, studied pea plants in his monastery’s garden to learn how traits such as color, shape, and size were passed down to the next generation. His work revealed some of the core principles of genetics and heredity.

Pham: This guy who really, you know, by himself, tried to understand something in the world that was a puzzle to him — and he just very meticulously worked through that. And his work wasn’t even discovered when he was alive. He didn’t get any recognition. But as the people rediscovered what he worked on, it’s basically changed genetics forever.

[“Dowdy” fades out]

Gearin: Thy loved learning about math and science but wasn’t sure what kind of career was the best fit for those interests.

Pham: We have very good science education in high school and middle school, but no one ever talked about doing research as a career. Like, we talk about it as something that you learn from textbooks, but there’s not this sense that, oh, these are questions that you can actually go and answer. It’s just like things that are presented to you, like a way of thinking about the world. And if people are good with science, I think that usually they go into become a doctor or an engineer. The way I read about science is mostly from scientists that work in isolation, like mathematicians. I didn’t have a sense that, oh, there’s institutions of people doing research. And there’s a platform, a framework for you to facilitate to do these things.

McElvery: But when Thy’s family made a big move, her opportunities shifted.

[“Eggs and Powder” begins]

Gearin: They came to Long Beach, California, where her dad’s siblings were living.

McElvery: Her parents hoped that living in the U.S. would provide more choices for Thy’s education. But the transition wasn’t without challenges.

Pham: We have a lot of extended family and, you know, all my uncles and aunts were taking very good care of us like, you know, guiding us through all the paperwork. But it was pretty different from where we were. I definitely was kind of scared. A lot of the culture aspects were very different from for me too. Like how people always smile when they see you, or ask you, ‘How are you doing?’ So I think that I was just pretty nervous about fitting in, and just trying my best to catch up with the language.

[“Eggs and Powder” fades out]

Gearin: Her knowledge of French, which shares common root words with English, helped her catch on, though.

Pham: I would definitely say that learning French for like 12 years helps me a lot with getting myself familiar with English.

McElvery: Enrolling at the nearby community college and taking English as a Second Language, or ESL, courses helped Thy gain momentum in American higher ed.

Pham: I took a bunch of ESL classes at Long Beach City College, and I’m just so very thankful for that opportunity because I think that ESL programs at the community college was really awesome. And I think my teachers were very caring, they paid a lot of attention to the students.

Gearin: Pop culture helped fill in the gaps in classroom language education.

Pham: Sometimes I joke that most of my spoken English actually comes from The Office because I watched it five times.

[“Net and Cradle” begins]

McElvery: While still learning English, Thy took on another challenge. As a junior, she transferred to the University of California, Berkeley as a premed student. While the curriculum was difficult, Thy’s persistence allowed her to study until the material made sense.

Pham: I remember my first semester at Berkeley, there was a class that was pretty challenging and we really had to pay attention in lecture. So what I did was I recorded the lecture and sometimes I would listen to it like two or three times and try to write out what I thought was essential.

Gearin: Her class on biophysical chemistry opened up a new way of blending her interests in biology and math.

Pham: Everything got built on a fundamental law of physics and chemistry.

McElvery: There was one experiment in particular that caught her attention.

Pham: It’s a biophysical experiment where you basically use two laser beams and then you can use the laser beam to grab on material. You just grab on one RNA molecule that forms a hairpin and then unfold it and see the kinetics of how it refolds. For me it’s really beautiful because it’s a direct way to measure what we expect the molecule’s behavior is. It was a way to think about biology very quantitatively, which is something I really like.

Gearin: She found a lab where she could explore those interests even further. She began working with Professor Carlos Bustamante, who studies the biophysics of individual biological molecules.

Pham: So being in the lab definitely made me seriously consider and finally decide to do research. I mean, research is hard, but I think people just go in every day ask different questions, talk to a lot of smart people, and discuss the problem they have and try to figure it out. That was something that was really appealing to me.

[“Net and Cradle” fades out]

McElvery: Following her curiosity, she found herself taking classes in a related field.

Pham: Computational biology was a natural transition there. Because Berkeley is very famous for computer science and at the time, when I would go to biology talks, a lot of people were talking about computational biology, you know, sequence analysis, alignment, and such. And I thought that was pretty neat. Yeah, taking computer science classes was a lot of fun.

Gearin: But it was getting time for her to decide what her next step would be.

Pham: It was the time where I was supposed to take the MCAT and I feel like, oh, actually, I don’t want to be a doctor.

McElvery: What she wanted to be was a research biologist.

[“Lissa” begins]

Pham: It would be something that matched my ability more, because I like to think about stuff, I like problem solving, and I like talking to people about ideas.

Gearin: She applied to a lot of different graduate programs but ended up trying to decide between Caltech and MIT.

McElvery: Thy opted for MIT Biology, because she was drawn to the computational biologists in the department. She was also curious about the climate.

Pham: I actually never really experienced a place with seasons before. So I thought that would be fun.

Gearin: Through her years as a grad student, she has grown to appreciate MIT Biology’s emphasis on training students to choose fruitful questions to study.

Pham: We all take classes in our first year and everyone here is an expert in a different field of biology. So I think that I got a lot of benefit from just talking to people about how they think about biology in general. I thought that, you know, scientists don’t need to talk to people at all, but I have realized that, you know, talking to people and discussing ideas is probably one of the most enjoyable parts of being a scientist.

[“Lissa” fades out]

McElvery: In her first year, she especially liked how one of her instructors, Whitehead Institute Member David Bartel, approached biological questions.

Pham: I just remember being very, very impressed with how smart and sharp he is in that paper reading class and how rigorous and careful he is as a scientist. I heard from people that, oh, like, you know, like Dave is always very available. You can always talk to him. And for me it’s like pretty rare, because like a lot of PIs are very busy with traveling. But Dave is always very available and always willing to mentor — talk to you about science.

[“Dance of Felt” begins]

Gearin: Thy joined Dave Bartel’s lab for her PhD project. There, she found a perfect opportunity to keep studying RNA while bringing the latest computational advances to bear on tiny, hard to predict molecules.

McElvery: The Bartel lab studies how gene expression in the form of RNA, the intermediary between the genetic code of DNA and the final products of proteins, is regulated by other molecules in the cell.

Gearin: RNA that codes for a protein is called a messenger RNA or mRNA. But cells need to fine-tune the amount of mRNA in the cell so that genes are expressed at just the right level.

McElvery: One of the ways cells do this is through a different type of RNA called microRNA. These short sequences bind to a protein called Argonaute or AGO that can target mRNAs to repress them, which means decreasing their levels in the cell. This repression ensures that genes aren’t overexpressed, which could lead to uncontrolled cell growth, and instead kept at a healthy level.

Gearin: Predicting which mRNA a given microRNA will target has proved to be a challenge, though. When a microRNA binds to AGO, the shape of the microRNA changes. That means you can’t simply read off the microRNA sequence and look for a complementary sequence in a target mRNA.

Raleigh: On top of that, a single microRNA can have many different mRNA targets, and they don’t bind each one with the same effectiveness. There’s a lot of variables in the system.

[“Dance of Felt” fades out]

Pham: You can, right now, predict 30 percent of variance. But at the end of the day, we still have 70 percent of the variance missing. So it means that there’s something about a system that we don’t understand enough to make good predictions.

Gearin: Thy wants to use deep learning methods to explain more of the variance — in other words, to make the lab’s predictions more accurate about what an individual microRNA does to decrease the levels of mRNA in the cell.

Pham: I think that if you really understand a system, then you’ll be able to make a very good prediction of how the system should behave.

McElvery: Two other graduate students, Kathy Lin and Sean McGeary, had done experiments where they precisely measured the mRNA targets for several microRNAs. That helped the lab understand the affinity between those microRNAs and their mRNA targets.

Gearin: Now, Thy wants to broaden those results to all microRNAs.

McElvery: An improved computer model could potentially give the researchers a way to predict, for any microRNA bound to an AGO protein, what mRNAs they seek out and how much they decrease the mRNA levels in the cell.

Pham: Basically, we want to have a biochemical model of how repression happens, and that model relies on the affinities between AGO microRNAs and the mRNA target. So what we wanted to do is we want to have to be able to predict local affinities for many, many microRNA target pairs.

McElvery: To predict the effects of any given microRNA, Thy is enlisting a new tool — neural networks. It’s a form of artificial intelligence that mimics the way that neurons in the human brain can forge unexpected connections.

[“Lakeside Path” begins]

Gearin:  It’s the same technique that has shown promise in complex tasks like colorizing black and white images and generating text in a chosen author’s style.

Pham: There’s a lot of fake Shakespeare going on the Internet nowadays, because a neural net was able to generate, like, perfectly fake Shakespeare.

McElvery: Instead of feeding her neural net Shakespearian plays, Thy will train it with known affinities between microRNAs and their mRNA targets that the lab has measured experimentally. Then the AI will look for more patterns than a human ever could, in order to build a better predictive model for all microRNA targeting.

Pham: The thing with the neural net is that it can try many, many functions in a very fast amount of time without you having to, like, supervise it. Some of them, as humans, we don’t even think of. So the neural net just has a way to explore a larger space than we are capable of.

Gearin: Thy hopes to finish collecting her training data for the neural net soon. She’s looking forward to delving into the computational part of the project.

Pham: I’m also very excited for it. I have to say, I probably like doing computational work more than experimental work, so I’m really looking forward to it.

[“Lakeside Paths” fades out]

I like to think quantitatively. I like to put like, numbers, symbols, pared-down everything in a very simple way. So for me that is appealing. For me at least, it’s not super different. It’s just a different way you answer a question.

McElvery: At the same time, Thy’s exploration of American culture continues. She volunteered as a canvasser during the 2020 primaries.

Pham: I learned actaully a lot from that. I don’t know a lot about American culture, so I’ve been also trying to, you know, get myself more familiar with the history and various aspects of American society.

[“New Crescente” begins]

Gearin: She has also contributed some time towards an initiative that mentors students at Bunker Hill Community College, since a community college helped her pivot towards her current career.

Pham: I think it’s like a great thing that our department is trying to do something like this. Because I came from community college, and I think that, you know, I didn’t really know about research back then. It’s like just a good way for you to really decide what you want to work out, like what your career should look like.

McElvery: She’s beginning to share her lessons about persistence with students interested in research.

Pham: I think that’s a lesson that, you know, every time my experiment fails I was like, oh God, it is so hard. But if you push at it, or talk to people and ask for help, you will eventually figure it out. Like it’s not — it doesn’t work like 100% of the time. But I think it’s given me the intuition to just keep trying.

[“New Crescente” fades out]

McElvery: Well, that’s all for today. Next time, tune in to hear about a student who is ­­­­investigating how cells can withstand and transmit mechanical force.

[“Something Elated” begins]

Gearin: Subscribe to the podcast on Soundcloud and iTunes or find us on our websites at MIT Biology and Whitehead Institute.

McElvery: Thanks for listening.

[“Something Elated” fades out]

Produced by Raleigh McElvery and Conor Gearin.

Music for this episode came from the Free Music Archive and Blue Dot Sessions at www.sessions.blue. In order of appearance:

• “Something Elated” — Broke for Free
• “Brek PKL” — Blue Dot Sessions
• “Dowdy” — Blue Dot Sessions
• “Eggs and Powder” — Blue Dot Sessions
• “Net and Cradle” — Blue Dot Sessions
• “Lissa” — Blue Dot Sessions
• “Dance of Felt” — Blue Dot Sessions
• “Lakeside Path” — Blue Dot Sessions
• “New Crescente” — Blue Dot Sessions