Attendees – Grace Peng, Susan Wright, Ben Dichter, Aaron Millstein, Christina Fang, Dana Greene-Schloesser, Doug Kim, Elizabeth Powell, Raj Bose, Yaki Stern, Shaul Druckmann, Susan Volman, Wolfgang Losert
First page of the wiki has all the projects
Behavioral data subgroup
Contact from the group – Dana and Susan V.
Unsure if the group is meeting…will ping the group to see if they are meeting
Can email them directly from the wiki
Had posted Adam’s presentation here but don’t see it linked so will have to check on that
Calcium imaging subgroup – Doug Kim is the contact Wolfgang Losert invited but don’t see him on the call but do see Ben Dichter on the call and hopefully he will endorse that we’ve had video conferencing calls. Wolfgang just joined. The calls were not recorded. Not sure if people were willing to share them but that’s something we can go back and ask for them.
Video conference activity once a month or every two months…talking about troubleshooting for calcium imaging. Trying to get all the users involved in the calls so everyone is aware of the advantages and disadvantages. Having another call next week…will be rescheduled for next month. Don’t remember who will be speaking. Would be great if everyone who is interested would participate for cross testing/validation.
Boot camp – might be misnomer – data intensive workshop. In June will be first U19 boot camp. Getting together and working together on Granger causality analysis. Will learn criticality analysis from a different expert on the team. Hidden agenda is that it gets the students and postdocs working together without the PIs setting the agendas. First white meeting in April…can tell you more then.
-Share downstream analysis with the group
-10 groups from different locations: Argentina, NY, Maryland, etc.
Wolfgang is now the lead and is passionate about trying to form a unified language for this.
*Grace encouraged everyone to update the wiki – can do password protected pages. Can post outcomes…a way to show off what you are doing. Sandbox page to take notes so we can remember when your next call and is and can designate someone in the group to take notes if you don’t have time. If you do designate someone, let us know and we will work with that person.
Modeling subgroup – Aaron and Shaul
Haven’t met since the last meeting
Meeting with the modeler would be useful
Grace uses their presentation a lot – they compiled the different models together…this is a great way to show what’s in the U19, who’s modeling what, where things can overlap, and where things can build on this to build more theories.
Shaul met with people separately at different conferences
For the groups doing modeling, hard to find ways for them to interact - they all have their own ways of doing things. Other things are coding practices. Aaron is developing a tool that could be useful to other groups because it’s general enough that other models implemented in Python could screen through hyperparameters. Different groups may have started that infrastructure. Could survey who’s going to cosin conference.
Talked about BRAIN 2.0 having bigger theories. NIH has a data science office and they have a program called the DATA Scholar Program. Wrote, in conjunction with BRAIN, one of these job postings. An opportunity to have a data scientist from industry/academia come work with us. Different positions and different ideas. One for BRAIN – touched on during PI meeting. This particular announcement is to bring someone in to do that using AI platforms and integrate, simulate, visualize, etc. the data. Came about after the last IMAG meeting that we hosted last year. Met a bunch of DARPA folks developing these tools. It would be great if you all could think about it through the context of building theories. Encourage you to advertise that as much as possible if you know of people who might be interested. Applications are due April 20th. Person would be on site in Bethesda…would become a federal employee for 1-2 years. We’re hoping we’d get someone, and that person would help you guys work with the U19s. For this particular job it’s a lot of curation of models. The person would help curate data, parameters. Implementation of research and implementation the latest models.
-One important consideration, especially for calcium imaging, is that the data and analysis are closely linked, so it’s hard to imaging unifying…what works for one may introduce artifacts into another model.
Raj’s group – have had some conversations – Michele is the NIH lead. Next Tuesday will be having another call – showcase – brief presentations on data platform. Data sharing, storage, various tools. For some people it’s been 2.5 years, for some people it’s been 1.5 years. Brought on Yaki Stern, who has a background in industry. Reaching a point where getting a little more sophisticated with discussion of tools Using Data Joint.
Looking at each of the tools and the categories it covers. Think the benefit is structuring the data.
-Can fill out forms to populate the wiki with the tools you are talking about. There’s also a credibility assessment page. You can link to and showcase these forms and convince people to use them. Go back to Activities page and then back to Other Interest Group.
Neuro Data Without Borders subgroup – Ben Dichter
Doing a lot of outreach – this group and others – labs. Archive – the whole data life cycle – the whole data life cycle – making sure the whole process makes sense all the way to archival. Tutorial scheduled for Cosyne Meeting – over 100 registrants. Within U19 – report progress on Calcium imaging calls – got feedback. A lot more work to be done. One thing we can do immediately is email people on list and tell them about Cosyne tutorial and post the slides and Jupyter notebook and if it’s recorded, even post that on this page. Going forward, having infrequent but consistent calls where we talk about progress or obstacles we are facing and how to organize this.
Overview of TMM section of Wiki – TMM creates tools for the neuroscience community, these tools will be disseminatable to the community. During BRAIN PI meeting in June, will be showcasing a lot of their tools – rotating group of slices. If some of the tools that you guys have are something that you want to showcase to have others use, you can share that with them. Will be creating pages for each TMM PI on the wiki.
PI meeting
Pre-meeting on May 31st – have room set aside for you all to meet. Had a call with the TMM PIs last week and they all wanted to have you meet with them.
Now we need to plan our time with you all and the U19s.
The TMM folks will probably do a flash presentation with you all in the room and then networking. What do you want to do?
Maybe not splitting people by subgroups because people are part of more than one subgroups but just having a U19 group. Just talking about practical areas, like data sharing, might be useful. Just one idea of not focusing on splitting everyone up by subgroups (at least not right away).
There is a half day booked – so potentially there is time to do something with the smaller groups and then everyone. (Most people think some subgroup time might still be good)
Brainstorming session – biggest bottlenecks – multimodal, multiscale registration – would be useful to revisit the notes from the past two meetings.
Joint group discussion - would be good to talk about any unifying challenges – serving own U19s or harmonizing between U19s
Opportunity for short methods talks? Something that hasn’t been public elsewhere. Tools that are further developed between U19s. Tools and methodologies.
Some subgroup time would be good if there is enough time
Tentative Agenda
May 31, 2020 pre-meeting with TMM and U19 Data Science Consortium
- TMM challenge discussion
- Flash presentations (include U19 methods being showcased)
- Networking
- Lunch
- U19 Methods Talks
- Subgroup presentations
- Joint group challenge discussion