Allen Institute’s Brain Observatory data set
Posted in Uncategorized
The Allen Institute has released the first set of data from their Brain Observatory project. Many of you already know about this, but I wanted to post about it to encourage people to take a bit of time to check out the data set themselves. They have a github page with materials that can help get you set up to explore and analyze their data.
Here’s why you should check it out: they’re listening. The team at the Allen Institute wants to create data that a broad audience would want to explore. Give them feedback and let them know what you’d like to see. This is a unique opportunity to influence a large-scale project. They’re really listening. Engage with them!
Has anyone managed to obtain raw images from the Allen Brain Observatory?
I followed their examples/instructions (install Anaconda, install allensdk, use their API to find the experiment ID, then download the data for a single “session”). The resulting .nwb data file contains meta-data and processed data (e.g. per-ROI fluorescence), but the acquired images are not in the .nwb file.
The more I read through the documentation, the more it seems like Allen are not releasing the acquired images, and only the post-processed data (after ROI selection). Any help or indication to the contrary would be welcome!
YES the data can and will be sent to you. email info@alleninstitute.org (link on the website for feedback) to inquire.
The ROI refer to the entire image area; it is all there.
I checked the dataset as well and came to the same conclusion (no raw files included).
It’s a pitty that they do not seem to release the raw images, but on the other hand this would mean a really large dataset which would be impractical. However, a couple of sample raw data would be interesting.
Hi, Same as for above: click the “send us a message” icon at the bottom of the main page so we know who is asking! We will send raw files if you want them but we would like to know a use case so we can make the site better. Your feedback is important- we are looking into posting “sample” datasets as an option for our next release, in late October.