Summer Workshop on the Dynamic Brain

October 5, 2015

Nestled in a pristine corner of the San Juan Islands on a peerless Pacific Northwest summer day, small groups of students huddle around picnic tables. The relaxed setting is punctuated by laptops and heated discussions, as students discuss how they will tackle the enormous data sets stored on innocuous-looking hard disks nearby.

The Summer Workshop on the Dynamic Brain, co-hosted every summer by the Allen Institute and the University of Washington at Friday Harbor Labs, is a singular course that brings together students and faculty from around the world to learn how to navigate the rich datasets produced by the Allen Institute.

The program includes lectures and hands-on workshops, and culminates in student-designed projects that use the data as a launch point to ask questions about the brain and how it functions.

“The workshop is useful for students, but it’s also a proving ground for our data and can actually influence how we package it in future releases,” says Lydia Ng, Senior Director of Technology. “It helps us get to know our audience by seeing what kinds of projects they choose to take on, what their expectations look like and how helpful the data is to them.”

This year, the workshop presented 24 students with several choices of data sets, including data and analysis tools for connectivity, cell types and the burgeoning Cortical Activity Map (CAM): a window into how vision works in real time in the brain.

“It’s incredible the amount of data they allow us to use,” says Joe Olson, a physics graduate student in a neuroscience lab at Harvard. “They’ve also already done a lot of the nitty gritty, first-level processing of the data so we can get right into interesting questions. That’s really amazing.”

For Olson and other students, the Friday Harbor workshop is a place to get ideas for Ph.D. projects and to learn how they can leverage Allen Brain Atlas datasets to make their questions deeper and more interesting.

“The workshop is great for getting not just an overview of the data but also overview of the field at large,” says Doris Voina, a graduate student at the University of Washington. “It’s general, interesting philosophical questions coupled with concrete aspects about the data and methods we can use to ask those questions.”

Students at the course are guided by faculty and teaching assistants, whose job it is to help students make the most of the data.

“Our role is to show students how to navigate the data,” says Saskia de Vries, Senior Scientist. “The CAM dataset consists of fluorescence traces, images, video and metadata, so we create pieces of code to help students pull just what they need.”

Students present their projects, individually and in groups, at the end of the two weeks. But the benefits of the workshop extend beyond the Friday Harbor Labs campus, for both the students and Allen Institute scientists.

“This workshop is by far the most hands-on of all the courses I’ve taken,” says Olson. “It’s already been enormously useful, and I can see how I’d use the Allen Institute resources for my Ph.D. work down the road.”

“The results of the workshop will help set the direction for the next data release,” says Ng. “Seeing students in action helps us determine what kinds of models and analysis we create and share with the larger scientific community.”