NeuroData Intro

class: center, middle

name:opening

# [Global Brain Workshop 2016](http://brainx.io)

.center[
presented by, Joshua T. Vogelstein
 
{[bme](http://www.bme.jhu.edu/),[icm](http://icm.jhu.edu/),[cis](http://cis.jhu.edu/),[idies](http://idies.jhu.edu/),kavli,[cs](http://engineering.jhu.edu/computer-science/), [ams](http://engineering.jhu.edu/ams/), [neuro](http://neuroscience.jhu.edu/)} | [jhu](https://www.jhu.edu/)
 
please ask questions: [jovo@neurodata.io](mailto:jovo at neurodata dot io)!
 
these slides: <http://brainx.io/ODEN>

]

---

# Details

- Date: April 6 & 7, 2016
- Host: Kavli Neuroscience Discovery Institute @ JHU
- Organizers: JT Vogelstein, R Huganir, and MI Miller
- Funders: NSF & Kavli

- 60+ scientists
- 15 "observers" (NIH, NSF, DARPA, IARPA, Simons, Kavli)
- 12 countries
- Report available from [brainx.io](http://brainx.io)

---

# Goals

Identify globally beneficial brain science projects:
- .w[Significant]: they will yield tangible  benefits to the world
- .w[Feasible]: we can achieve major milestones given existing funding 
- .w[Inclusive]: all nations can meaningfully contribute  and benefit
- .w[Collectively diverse]:  span a wide range of brain sciences

---

## 3 Challenges...

- What makes our brains unique?
- How does the brain solve the complex  problems of intelligence?
- How can we augment mental health decisions with data?

##  and 1 Universal Resource...

<!--

<center><Big>The International Brain Station (TIBS)</Big></center>
 -->

---
class: center

## The International Brain Station (TIBS)

---

## Motivating Themes

- Eliminate "pain point" for each stage of discovery
- Open science (data and tools)
- Extensive collaboration
- Reference research artifacts
- Continuous integration and agile development
- Cultural sensitivities

---

# Collect

.pull-left[
- Pain-point: Locally "baby-sitting" hardware
- Solution:  *Data Collection Dashboards*

- Interface directly with measurement machines
- Live status updates and summaries
- Enables remote intervention
- Automatic metadata generation
]

.pull-right[
<figure>
<img src="images/deluge.jpg" alt="Drawing" style="width: 300px;"/>
 <figcaption>source:economist.com</figcaption>
</figure>
]

---

# Store

.pull-left[
- Pain-point: Data too large to store/analyze locally
- Solution:  *Multimodal Cloud Storage*

- Interface directly with collection machines
- Organizes data for efficient queries
]

.pull-right[
<figure>
<img src="images/sejnowski.jpg" alt="Drawing" style="width: 300px;"/>
 <figcaption>source:nature.com</figcaption>
</figure>
]

---

# Explore

.pull-left[
- Pain-point: Want to see data from anywhere
- Solution:  *Interactive Data Explorers*

- Mobile compliant (1 billion+ additional users)
- Augmented reality environments
]

.pull-right[
<figure>
<img src="images/explore.jpg" alt="Drawing" style="width: 300px;"/>
 <figcaption>source:ultravr.org</figcaption>
</figure>
]

---

# Analyze

.pull-left[
- Pain-point: Raw data are messy and stored differently
- Solution: *Reference Analysis Pipelines*

- Streaming analysis
- Continuously integrated with new data & methods
- Notebooks for easy replication and extension
]

.pull-right[
<figure>
<img src="images/jupyter.png" alt="Drawing" style="width: 300px;"/>
 <figcaption>source:treasuredata.com</figcaption>
</figure>
]

---

# Model

.pull-left[
- Pain-point: Model and data are incommensurate
- Solution: *Data Modeling Bots*

- Automatically fit new models to old data, and new data to old models
- Community driven metrics, benchmarks, and leaderboards
]

.pull-right[
<figure>
<img src="images/hubel1.jpg" alt="Drawing" style="width: 300px;"/>
<img src="images/hubel2.png" alt="Drawing" style="width: 300px;"/> 
<figcaption>source:wordpress.com</figcaption>
</figure>
]

---

# Design

.pull-left[
- Pain-point: Extensive simulations
- Solution:  *Synthetic Data Simulation Environment*

- Extracts simulation parameters from datasets
- Runs virtual experiments
]

.pull-right[
<figure>
<img src="images/tea.jpg" alt="Drawing" style="width: 300px;"/>
 <figcaption>source:lady tasting tea</figcaption>
</figure>
]

---

# Search

.pull-left[
- Pain-point: Data are scattered and poorly labeled
- Solution: *Metadata Query Service*

- All data  and code will have community driven metadata
- Queries to support multiple languages
]

.pull-right[
<figure>
<img src="images/search.jpg" alt="Drawing" style="width: 300px;"/>
 <figcaption>source:Minority Report</figcaption>
</figure>
]

---

# Educate

.pull-left[
- Pain-point: New research platform
- Solution: *Extensive educational focus*

- Culturally sensitive MOOCs
- Quarterly training workshops
- Wiki-style documentation and tutorials
]

.pull-right[
<figure>
<img src="images/educate.jpg" alt="Drawing" style="width: 300px;"/>
 <figcaption>source:wikipedia.org</figcaption>
</figure>
]

---

## Challenges

- Cultural shift in neuroscience
- But cosmology and molecular genetics did it

---

## Implications for ODEN

- .w[Incentives]: lowers the barrier to entry into each stage of analysis
- .w[Sustainability]: resources (data & code) can trivially be supported by communities who use them
- .w[Discoverability]: all resources will, by default, include extensive metadata

---
class:   middle, center

# Questions?

e: [jovo@neurodata.io](mailto:jovo@neurodata.io)

w: [brainx.io](http://brainx.io)