Predicting gene expression with AI

When the Human Genome Project succeeded in mapping the DNA sequence of the human genome, the international research community were excited by the opportunity to better understand the genetic instructions that influence human health and development. DNA carries the genetic information that determines everything from eye colour to susceptibility to certain diseases and disorders. The roughly 20,000 sections of DNA in the human body known as genes contain instructions about the amino acid sequence of proteins, which perform numerous essential functions in our cells. Yet these genes make up less than 2% of the genome. The remaining base pairs — which account for 98% of the 3 billion “letters” in the genome — are called “non-coding” and contain less well-understood instructions about when and where genes should be produced or expressed in the human body. At DeepMind, we believe that AI can unlock a deeper understanding of such complex domains, accelerating scientific progress and offering potential benefits to human health.Read More

From intern to FTE: Four researchers share their journeys with the Facebook Core Data Science team

Facebook’s Core Data Science (CDS) team is pushing the envelope of what’s possible by exploring and solving novel challenges in science and technology. The CDS internship program was designed to provide researchers with the opportunity to explore career paths in diverse fields of research, such as computer science, statistics, optimization, economics, and social science.

CDS interns are immersed in Facebook culture and see the direct impact their research has on product development, user experience, and the research community as a whole. When transitioning to full-time employees, researchers use passion and curiosity to drive their projects. Here, four researchers share their experiences with the program and what inspired them to transition to their full-time roles.

Staying curious while embracing an exploratory mindset

Taylor Brown is a computational social scientist on the CDS team. She has a PhD in sociology from Duke University, where she specialized in culture and computational methods and networks. She also holds an MSc in evidence-based social intervention from the University of Oxford. Taylor develops methods to improve measurement and mitigation of social inequalities and harmful content on Facebook and Instagram while protecting user privacy.

“As part of my CDS internship at Facebook, I was able to join a groundbreaking collaboration with academic researchers to study the role social media played in the 2020 U.S. presidential election,” says Taylor. “All studies in this collaboration are preregistered, and now, as a full-time team member, I am contributing to their publication. This work will no doubt impact many fields of social science for years to come.”

Taylor’s curiosity has driven her research career and led her to a full-time role at Facebook after two attempts to apply for a CDS internship. “I didn’t make it through the interview process when I first applied for an internship at Facebook,” she says. “The experience taught me a lot about Facebook’s culture. I felt supported all the way through, and I was encouraged to apply again when the next round of positions opened. I’ve now met half a dozen CDS team members who also needed to apply twice to be accepted into the intern program. After having this experience, I always tell people to try again if they don’t get through the interview process the first time around.”

Within CDS, team members have autonomy to focus on projects they feel passionately about. For Taylor, it was easier to transition into a full-time role after two internships. “I recommend completing two internships if you can,” she offers. “I used my first internship to understand Facebook’s culture, research tools, and processes. This set me up for success in my second internship, in which I was able to jump right into project work with the team.”

When deciding whether to stay in academia or move into full-time industry research, it’s important to consider what you want from your career and lifestyle. Taylor found that she appreciated the work/life flexibility Facebook offered and that her internship experience debunked some ideas she had about work outside of academia. “I was afraid industry research would be less stimulating than the free thought of academia, but it’s not,” she says. “There’s no shortage of important questions to answer or complex challenges to solve, and I’m constantly enriched by my coworkers, all of whom are incredibly talented scientists and engineers.”

Discovering new perspectives and keeping an open mind

Yanyi Song earned her PhD from the University of Michigan in Ann Arbor and focuses on Bayesian methods, causal inference, and statistical learning. Yanyi transitioned into a full-time research scientist position last year after completing a data science internship. She saw an online post about the internship program and felt the job was an ideal match for her skills, interests, and career aspirations.

During the program, Yanyi worked on optimizing the user experience with Facebook Ads Manager. “I was motivated by the challenge to think from the user’s perspective, which was new for me,” Yanyi says. “We were tasked with building a data-driven model to improve the user experience, and it took a lot of cross-functional collaboration to reach our goals. As an intern, you have access to exceptional researchers who are incredibly supportive. They’re there to help and answer questions and empower you to share your ideas.”

“I also learned how important communication is when working with different departments,” she continues. “CDS works with teams across Facebook, and it’s a much different environment and pace than academia. I recommend anyone interested in joining CDS prepare themselves to ask questions and maintain a growth mindset. At CDS, you get to work on projects you’re passionate about, and there are endless opportunities to learn and try new things.”

Developing new skills and solving complex challenges

Daniel Thomas recently completed a CDS internship and is now a research scientist at Facebook. He holds a PhD in political science from Columbia University, where he studied the dissolution, formation, and reformation of social networks during and after conflict.

“When interning with the CDS team, there are set parameters for your work,” Daniel says. “My internship expanded on my areas of interest and challenged the skills I’d developed during my PhD program. I had the opportunity to work on a project designed to help us evaluate new network disruptions that the Engineering team was implementing on Facebook. I was surprised by the fast-paced environment and how quickly we can make an impact here compared to the research, review, and publishing process in academia. Interns play an important role on teams and are expected to present ideas, recommendations, and share code and early numbers in meetings. Though it was difficult for me to feel OK with bringing unfinished results to these meetings at first, the challenge empowered me to quickly evolve the efficiency of my coding.”

Daniel was inspired by the scalable impact he was making at Facebook and became a full-time member of the CDS team after his internship ended. “The biggest difference I’ve noticed after my transition to a full-time role is the strong sense of ownership you’re given,” he says. “Rather than being assigned specific tasks, you’re responsible for defining your own projects and moving them forward. Facebook gives you a lot of support while you create your research path, which is incredible. I’ve learned to be more proactive and to overcommunicate. Don’t be afraid to ask questions, and set clear expectations to help you manage your time.”

Feeling inspired to make a meaningful impact

A graduate of the University of Minnesota Twin Cities, Saurabh Verma has a PhD and a master’s degree in computer science and is a research scientist on the CDS team. His research focus is in machine learning and covers deep learning models, natural language processing for Q&A systems and machine reading comprehension models, graph neural networks for graph representation learning, knowledge graphs for information retrieval, and social network analysis.

Saurabh wasn’t looking for an internship when he found a passionate team of problem solvers from diverse backgrounds at CDS. “I was presenting a paper at a conference a few years ago and was invited to a Facebook networking event, where I met my future manager,” he says. “I found the interview process really inspiring. My conversations with recruiters and managers revealed vast opportunities to solve novel challenges, including my work on integrity management and abusive account detection during my two internships.”

Researchers can continue in the CDS intern program if they are still working on their PhD when they finish their first three-month internship. “I received a full-time offer after I graduated and completed my second internship last year,” Saurabh says. “As an intern, I was hyperfocused on specific sections of a big project. I’ve enjoyed being challenged to see the broader project vision, from start to finish, in my full-time role.”

He continues, “My advice to anyone considering a CDS internship — and ultimately a full-time research position at Facebook — is to think hard about the work you see yourself doing. Ask yourself: ‘What projects keep me inspired?’ You never know when opportunities will present themselves, but knowing which projects you’re passionate about may help you stumble into something great.”

—

Are you interested in applying for an internship with the CDS team? Check out the careers page.

The post From intern to FTE: Four researchers share their journeys with the Facebook Core Data Science team appeared first on Facebook Research.

Read More