Neurohackweek 2017 is a 5-day hands-on workshop in neuroimaging and data science, held at the University of Washington eScience Institute, September 4th-8th, 2017.
Participants will learn about technologies used to analyze human neuroscience data and develop the skills needed to make their analyses and results shareable, transparent, and reproducible.
Welcome to Neurohackweek 2017! Neurohackweek 2017 is an NIH and Moore Foundation-supported 5-day hands-on workshop in neuroimaging and data science held at the University of Washington eScience Institute, September 4th-8th, 2017. Our goal is to help researchers learn about technologies and tools at the interface between human neuroimaging and data science, and develop the skills needed to make analyses and results shareable, transparent, and reproducible. Neurohackweek is modeled on the successful Astrohackweek annual workshop also based at the eScience Institute.
Neurohackweek invites approximately 40 early-career researchers (mostly senior graduate students and postdocs or other junior PhDs) with backgrounds in neuroimaging and/or computational science to spend a week in Seattle collaborating with each other on a range of projects. At the 2017 installment of Neurohackweek, we'll have participants from 8 different countries around the world. Admission is competitive, and in 2016 and 2017, we've unfortunately only able to invite 25% - 30% of applicants. (Thanks to new funding from NIMH, we anticipate increasing attendance at Neurohack 2018 to at least 60 participants.)
Neurohackweek is directed by Ariel Rokem and Tal Yarkoni, and is supported by an R25 award from the National Institute of Mental Health as well as by the Moore-Sloan Data Science Environments initiative.
We are fortunate to have a stellar rotating cast of instructors at Neurohackweek. Our course faculty have expertise in multiple modalities of human neuroimaging and a wide range of areas of data science and informatics. For the 2017 installment of Neurohackweek, faculty include:
- Bing Brunton (University of Washington)
- Nicholas Cain (Allen Institute for Brain Sciences)
- Satra Ghosh (MIT)
- Chris Gorgolewski (Stanford University)
- Bernease Herman (UW eScience)
- Chris Holdgraf (University of California, Berkeley)
- Anisha Keshavan (University of Washington)
- Tara Madhyastha (University of Washington)
- Jeanette Mumford (University of Wisconsin, Madison)
- Dylan Nielson (NIMH)
- Russ Poldrack Stanford University
- JB Poline (McGill University / MNI)
- Ariel Rokem (UW eScience)
- Valentina Staneva (UW eScience)
- Jake Vanderplas (UW eScience)
- Tal Yarkoni (University of Texas at Austin)
We've done our best to streamline the course and ensure that people can hit the ground running on Day 1. However, to ensure that everyone has a good experience both professionally and personally, we'd like everyone to do two things at the outset of Neurohackweek--actually, preferably ahead of the course. The first thing is to install as much of the course software as possible ahead of time, and the second is to read--and take seriously--our Code of Conduct.
We strongly recommend that participants spend some time ahead of the course making sure they have most of the software they'll use during NHW installed. Because of our tight schedule, we're not able to set aside much time on Day 1 for installation; aside from a few minutes between tutorials, we're going to assume you already have all of the software discussed by the instructors installed. Unfortunately, we also can't provide laptops or desktop computers to participants--so you'll need to come with your own laptop. If you absolutely cannot find a suitable computer to use for the course, please contact Ariel Rokem, and we'll do our best to find you one you can use.
To maximally benefit from Neurohackweek, participants should install the software packages and tools described (with installation instructions for all major operating systems) on this page. We encourage you to do this ahead of the course, so that you can spend your time on Day 1 following the tutorials rather than reading installation instructions and missing what the instructors are saying.
Though we suggest installing everything on the provided list, the most critical elements--without which you're probably going to have a bad time at Neurohackweek--are the bash shell (which is already built into the operating system if you're on Linux or OS X), git, and a good text editor (or, if you're fancy, a full-blown IDE). Additionally, because most of the instruction in the course will be conducted in Python, we also strongly encourage everyone to have a working Python environment. To round things out, we suggest installing RStudio and Docker, which will be used for various tutorials and demonstrations.
In addition to an installation of the software on your own machine, and in case there is any software that you did not manage to install, we also prepared a deployment of Jupyterhub that has much of the software already installed. You can log into this deployment with your Github user account at: https://neurohackweek.github.io/jupyterhub.
Neurohackweek is dedicated to providing a harassment-free learning experience for everyone regardless of gender, gender identity and expression, sexual orientation, disability, physical appearance, body size, race, age or religion. We do not tolerate harassment of participants in any form. Sexual language and imagery is not appropriate for any conference venue, including talks. Conference participants violating these rules may be sanctioned or expelled from the conference without a refund at the discretion of the organizers.
Harassment includes, but is not limited to:
- Verbal comments that reinforce social structures of domination related to gender, gender identity and expression, sexual orientation, disability, physical appearance, body size, race, age, religion.
- Sexual images in public spaces
- Deliberate intimidation, stalking, or following
- Harassing photography or recording
- Sustained disruption of talks or other events
- Inappropriate physical contact
- Unwelcome sexual attention
- Advocating for, or encouraging, any of the above behaviour
Participants asked to stop any harassing behavior are expected to comply immediately. If a participant engages in harassing behaviour, the organisers retain the right to take any actions to keep the event a welcoming environment for all participants. This includes warning the offender or expulsion from Neurohackweek with no refund.
Neurohackweek organisers may take action to redress anything designed to, or with the clear impact of, disrupting the event or making the environment hostile for any participants. We expect participants to follow these rules at all the event venues and event-related social activities.
Harassment and other code of conduct violations reduce the value of Neurohackweek for everyone. If someone makes you or anyone else feel unsafe or unwelcome, please report it as soon as possible to one of the instructors. You can make a report either personally or anonymously. Anonymous reports can be made at https://neurohackweek.wufoo.com/forms/neurohackweek-code-of-conduct-violation/
-- This anti-harassment policy is based on the example policy from the Geek Feminism wiki, created by the Ada Initiative and other volunteers.
Aside from the nuts and bolts of the course, we've also tried to consolidate in one place (specifically, this place!) a few other informal policies and suggestions that we think NHW participants may benefit from keeping in mind.
First, Neurohackweek is first and foremost about interaction and collaboration. It places much more emphasis than most other training courses on hands-on collaboration and extended interactions between participants (and instructors). This bears emphasizing, because in our experience, no matter how many times we say it, many participants don't take enough advantage of the interactive format of the course. Both the instructors and your fellow participants at NHW are here for you. As such, you should feel free to ask anyone else attending NHW any questions that you think might fall within their areas of expertise, to bounce ideas off of them, and to solicit feedback and input on your team projects as they evolve. Or even just ask them what they're working on, out of curiosity. The point of the instructors and other participants being here, and of NHW adopting a hands-on format, is to help you acquire new information and new skills; if you think someone else can help facilitate that, go talk to them!
Concretely, the rule we'll try to adhere to at NHW is that if someone's hanging around in a common area (e.g., the DSS), they're fair game for interaction. This cuts both ways, of course: your fellow participants (as well as the instructors) are also going to feel comfortable asking you for your help or input when they need it. Of course, this doesn't mean you have to prioritize others' needs above your own work; if you're concentrating intently on something and don't want to be interrupted for a while, that's totally cool--just duck into a quiet corner somewhere (there will be break-out rooms people can use in the Data Science Studio in the afternoons), or let whoever's asking you for help know that you'll come find them once you're done.
There's one important caveat to the "interactivity" rule above that we need to mention: many of the people who'll be working in the Data Science Studio while Neurohackweek is going on are full-time employees or researchers at the eScience Institute, and aren't there just for the course. This includes the eScience-based NHW instructors (Bernease, Jake, and Valentina), who have been kind enough to volunteer their time to help with the tutorials, but have plenty of other non-NHW-related duties to attend to otherwise. Please be respectful of these people's time and give them some space. Keep in mind that we're essentially holding a 50-person week-long course in the middle of their workspace. :)
In general, most of the tutorials and break-out sessions at NHW will be conducted in the Python programming language, to the relative exclusion of other languages used in the neuroimaging community (though there will be tutorials in R and JavaScript, and probably break-out sessions that use other languages). We want to emphasize that the decision to focus primarily on Python is purely a pragmatic one: it would be very inefficient, in an intensive week-long course that emphasizes development of software development and data science skills, to try to teach and work in several different languages. By emphasizing a single language, our hope is that participants will be able to focus on learning concrete computational and data science skills, rather than spending their time learning the idiosyncracies of each new language's syntax.
As we'll discuss on Day 1, we also happen to think that Python is a particularly good choice for our course, because (a) it's relatively easy to learn; (b) it's a general-purpose language that has excellent tooling support in a vast number of domains; (c) it's already very widely used in the neuroimaging data science community; and (d) it's completely free and runs on all major operating systems. That said, we want to be very clear that the fact that Python will be the main language of instruction at NHW has no implications at all for what languages you can and should use during the course. A huge amount of neuroimaging software is written in languages other than Python, and there are many, many neuroimaging applications where the best-in-class tools are written in MATLAB, R, C, or something else. While we encourage all participants who aren't familiar with Python to spend a bit of time working in it, we're happy for you to work in whatever language you like. As a point of reference, in the inaugural installment of NHW, participants contributed to projects written in Python, R, MATLAB, JavaScript, and C (as well as some projects written only in natural languages!).
Neurohackweek is, in at least this one respect, not like Vegas: what happens at Neurohackweek doesn't stay at Neurohackweek; instead, it gets shared publicly with the rest of the world. All of the materials from all of the tutorials, break-out sessions, project presentations, etc.--as well as videos of all of these presentations--will be posted publicly (ideally, in real-time). We will also be taking photos of the goings-on at Neurohackweek fairly regularly (and you're welcome to as well). We recognize that some people may have reservations about being included in media disseminated on the web, and if you have such concerns, please come and talk to Ariel Rokem or Tal Yarkoni. We will then do our best to try and extirpate you from the record. Please recognize, however, that sharing course materials, tutorials and projects with the rest of the neuroimaging and data science communities is an explicit goal of Neurohackweek. So while we'll gladly acquiesce to any request not to have photos containing one's visage displayed pubicly on the @Neurohackweek Twitter account, we do expect all participants to do their best to publicly share all of the code, text, or other materials they generate during Neuorhackweek (and we will of course be providing lots of instructions and help intended to facilitate that sharing process).
The NHW17 schedule consists of several different kinds of events: data science tutorials; traditional lectures; hacking/team project sessions; and break-out sessions. Oh, and meals. Meals are also a kind of event that we will have, because in past years we've found that Neurohackweek participants learn and contribute more when they're not suffering from starvation.
All events on the schedule will be held in one of four locations: the Lander Hall (LH) residence where all participants and instructors are staying; the Data Science Studio (DSS) on the 6th floor of the Physics-Astronomy Building; the Odegaarde Undergraduate Library (OUGL); and the classrooms on the 5th floor of the Physics-Astronomy Building (PAB5). Locations for all events are indicated in parentheses below using the above abbreviations. The general idea is that breakfast and lunch will always be in Lander Hall, and (at least after Monday) morning sessions will be at the Odegaarde Library while afternoon sessions will be in the Data Science Studio. All of these locations are an easy (and very pleasant) 10-minute walk from one another. Note that large groups of participants and instructors will be walking together between these locations, so if your navigational skills are not so great, not to worry--just follow one of the people around you who look like they know where they're going (because that never ends badly).
We'll be updating this document continuously as NHW17 unfolds (and you'll be expected to too!), so links to slides, video, etc. will be added to the schedule below as they become available.
Arrival and check in. All participants and instructors will be staying in the Lander Hall dorm adjacent to the main UW campus (the full address is Lander Hall, 1201 NE Campus Pkwy, Seattle, WA 98105). To get to Lander from the SeaTac Airport, you have two main options:
-
The cheaper and slower way is via public transportation, which will take about an hour. From the arrival area, follow signs that say "to Link light rail". You'll walk 5 - 10 minutes to the SeaTac Link stop. There, buy a one-way ticket to the University of Washington stop (which is the last stop in the direction of downtown), which will cost $3.25. When you get to the UW Station (do not get off at University Station, which is downtown and not at the University of Washington!), exit the station and walk approximately 15 minutes to Lander Hall (you can also take one of several buses if you have large bags, but it won't save you much time, and the walk is quite pleasant if you cut through campus). A more detailed Google Maps map can be found here.
-
Alternatively, you can take a taxi or ride-share directly from the arrival area at the airport to Lander Hall. Barring serious traffic (unlikely on a Sunday), the ride should take 20 - 25 minutes, and will probably cost around $35-$50.
You can check into your room at Lander Hall at any time after 2 pm on Sunday, September 3rd (though you can temporarily store your bags at the reception if you arrive earlier than that). Upon check-in, you'll receive a room key as well as a conference card that will grant you access to the dining hall in Lander for breakfast and lunch all week (you are responsible for your own dinner on Tuesday, Wednesday and Thursday). Please be sure to check out by 11 am on Saturday, September 9th (you can store your luggage at the desk for the day if you'd like).
The schedule on Monday is a little different from the rest of the week, in that the afternoon will consist of data science tutorials rather than team project/hacking sessions. The idea is to make sure that everyone has the basic skills needed in order to jump in and contribute to the group projects the rest of the week.
-
7 AM - 8:35 AM: Breakfast (LH)
-
8:35 AM: Meet at the Lander check-in desk to walk over together to the Physics-Astronomy Building. Note that the building will be locked for the Labor Day holiday, so please be at the reception at 8:35 AM (otherwise you'll have to ask someone on Slack to come down and let you in).
-
9 AM - 10:30 AM: Introduction to Neurohackweek (Ariel Rokem; PAB5)
-
10:30 AM - 12 PM: Reproducibility in fMRI: What is the problem? (Russ Poldrack; PAB5)
-
12 PM - 1 PM: Lunch (LH)
-
1 PM - 6 PM : Data science tutorials (PAB5)
-
7 PM : Welcome reception (DSS)
On Monday afternoon, we will have two parallel tracks of data science tutorials, held simultaneously in different rooms. You're welcome to move between rooms as you like. We've tried to organize the tutorials in such a way as to maximize the differences between the topics being covered at any given time, but it's inevitable that some people will want to attend both of the tutorials being held at a given time. While the instructor cloning machine we purchased last year is currently out of service, the good news is that video recordings and materials for all of the tutorials will be immediately available online, so you'll be able to catch out on what you missed in short order if you so choose. Also, most of the tutorial instructors on Day 1 will be hanging around for most or all of the rest of the week, and you should feel very free to approach them if you have any questions about anything they covered (or for that matter, about related topics that they didn't cover).
| Time | Room 1 | Room 2 |
|---|---|---|
| 1 - 2 PM | Unix shell 1 (Valentina) | Best practices in scientific computing (Chris): Software containers (Docker) |
| 2 - 3 PM | Unix shell 1 (Valentina) | Best practices in scientific computing (Chris): Software testing |
| 3 - 4 PM | Git/GitHub 1 (Bernease) | Intro to data viz with D3.js (Anisha) |
| 4 - 5 PM | Git/GitHub 2 (Bernease) | Jupyter/Interactive computing (Dylan) |
| 5 - 6 PM | Python programming 1 (Tal) | Intro to R (Jeanette) |
| 6 - 7 PM | Python programming 2 (Tal) | Intro to R (Jeanette) |
For the days Tuesday through Thursday, we'll adopt a fairly fixed schedule: we'll be at Odegaarde Library in the morning, and the DSS in the afternoon. The "official" events will wind down around 6 pm every day, but we've reserved a nice space in the Lander Hall dorm that will be available until 12 am every night, and you're welcome to keep hacking there with other participants if you're so inclined. (If you're not so inclined, that's completely fine--we recognize that 5 days of 9-to-6 scheduling is already pretty intense).
Remember that no dinner is provided on Tuesday through Thursday, so you're on your own for those days. We encourage you to explore the neighborhoods around campus and/or other parts of the city--Seattle has fantastic food options at all budget levels.
7 AM - 9 AM: Breakfast (LH)
9 AM - 12 PM: Cloud computing (Tara Madhyastha; OUGL)
12 PM - 1 PM: Lunch (LH)
1 PM - 3 PM: Team project pitches (PAB5)
3 PM - 6 PM: Team projects (DSS)
- Afternoon breakout sessions (optional):
- 3 PM: AIBS data-sets (Nick Cain)
- 4 PM: Advanced data manipulation in pandas (Tal Yarkoni)
6 PM - 12 AM: Dinner (on your own) and hackathon sessions (LH)
7 AM - 9 AM: Breakfast (LH)
9 AM - 11 AM: Machine learning (Jake Vanderplas; OUGL)
11 AM - 12 PM: Machine learning in neuroimaging (Chris Holdgraf; OUGL)
12 PM - 1 PM: Lunch (LH)
1 PM - 6 PM: Team projects (DSS)
- Afternoon breakout sessions (optional):
- 3 PM: Statistical practice in neuroimaging (JB Poline)
- 4 PM: Efficiency for fMRI (Jeanette Mumford)
6 PM - 12 AM: Dinner (on your own) and hackathon sessions (LH)
7 AM - 9 AM: Breakfast (LH)
9 AM - 12 PM: Neuroimaging pipelines (Satra Ghosh; OUGL)
12 PM - 1 PM: Lunch (LH)
3 - 6 PM: Team projects (DSS)
- Afternoon breakout sessions (optional):
- 3 PM: Advanced time-series analysis (Bing Brunton)
6 PM - 12 AM: Dinner (on your own) and hackathon sessions (LH)
7 AM - 9 AM: Breakfast (LH)
9 AM - 12 PM: Team projects (DSS)
12 PM - 1 PM: Lunch (LH)
1 PM - 4:00 PM: Project presentations (PAB5)
4:00 PM - 5:00 PM: Summary and feedback (PAB5)
6:00 PM: Meet at AGC Marina to board cruise boat
6:30 PM: Farewell reception and Lake Union Cruise
On the last night of Neurohackweek, we'll have a farewell reception aboard a cruise on Lake Union. We have to be at the AGC Marina at 6:00 PM to board our cruise boat; sailing time is 6:30 PM sharp (anyone not on board by 6:30 PM gets left behind, sadly).
Getting to the AGC Marina by public transport is straightforward. Bus line 70 picks up on 15th Ave and Campus Parkway, and drops off at SLU, a short walk from the AGC marina. The trip takes around half an hour (but allow for traffic when planning!), and one-way fare is $2.50. Alternatively, you can take a car-share or taxi to the Marina. Depending on traffic, the car ride takes anywhere from 12 to 25 minutes, and should cost $10 - $20.
Check out and departure. Sadly, all good things must come to an end. In this case, the end must come by 11 am on Saturday, September 9th--when is the latest you can check out of Lander Hall.
This is a text mining project that uses automatic web scraping of relevant literature given terms of interest from the Cognitive Atlas (Poldrack et al., 2011). It includes a few tools from natural language processing to analyse and visualize the resulting corpus of literature.
Contributors: Tom Donoghue, Ayala Allon, Basak Kilic, Eric Reavis, Mengya Xhiang, Juliet Shafgo, Nicole Roberts, Vassiki Chauhan
https://github.com/neurohackweek/DataDrivenCognitiveOntology
(Team members: Shruti, Scott, Suyash, Dora) Code: its-going-to-be-there, along with plots!
https://github.com/neurohackweek/avalanche.
(team members: Anisha, Dylan, Kesshi, Alejandro).
Link to slides: https://drive.google.com/open?id=1Mik0V0C5EPATg4ls1E0jgyBHqN8WFZL3wmTic Code: https://github.com/medulina
Contributors: @mikhail, @dewarren
Code: https://github.com/mibel/mri_segmentation_dl
Using machine learning to predict individual differences(team members: Patrick, Dani, Kevin, Jeanette, Sumedha, Gary, Thomas, Julia).
Code: https://github.com/neurohackweek/HCP_MachineLearning Slides: https://docs.google.com/presentation/d/1inSH-0V1oWEwX9v53SKM8NBtVxS9mpjhJbbYdKeaXc4/edit#slide=id.g260932d089_4_9
Team members: Satra + Chris + a ton of other people
neuro-data-science/neuro_data_science#14 https://github.com/neuro-data-science/neuroML
AWSome stuff, but not really, it’s more like rAWSignal or nAWSombody else should figure out AWS Lambda
Team: Daniel Moyer
Code: https://github.com/dcmoyer/boto_bench
Team members: Chris, Tal + Sumedha Code: https://github.com/chrisfilo/peaks2maps
3D surface visualizations as well as interactive 3D volume and surface visualizations.
Team members: Murat Bilgel, Benjamin Ely, Melanie Ganz, Krisanne Litinas, and Andrea Shafer, heavily indebted to Satra Ghosh, Chris Holdgraf, Anisha Keshavan, and Tal Yarkoni
Code: https://github.com/neurohackweek/NHW_visualization and https://github.com/neurohackweek/niwidgets
Team members: Derek, Grace, James, Aki, Josh
Code https://github.com/dpisner453/PyNets
Neurohackweek is supported through a grant from the National Institute for Mental Health, and through a grant from the Gordon and Betty Moore Foundation and the Alfred P. Sloan Foundation to the University of Washington eScience Institute Data Science Environment.
We are also grateful for a close collaboration with the Jupyterhub team and for their help in setting up our deployment of Jupyterhub.