Recap of COSYNE 2018

The Computational and Systems Neuroscience (COSYNE) conference provides a forum for the exchange of experimental and theoretical/computational approaches to problems in systems neuroscience.

The conference is divided into two main sessions:

  • Main meeting (Denver, CO, 1-4 March 2018)
  • Workshops (Breckenridge, CO, 5-6 March 2018)

Tutorial Session

This year, COSYNE was hosting its first tutorial session, “Statistical models for neural data: from GLMs to latent variables”, held by Prof. Jonathan Pillow (Princeton University). There were 293 registered for the tutorial, and it was their first COSYNE for [155, 208] (95% CI) participants. Slides can be found on his website.

Main Meeting

The main meeting consists of a single track featuring both invited and submitted (high-scoring abstracts) talks. This year saw another record in the number of attendees and submitted abstracts (857 registrations, 709 abstracts submitted, 396 accepted - 55.8%), which was increased from 330 accepeted abstracts last year in Salt Lake City, UT thanks to a bigger venue in Denver.

Lots of presentations were dominated by a dynamical systems view of the brain. More and more researchers are taking the view that whereas individual neuronal responses are very high-dimensional, the activity of a population of neurons is usually restricted to a lower-dimensional manifold (i.e., a topological space that locally resembles Euclidean space near each point) within the space of all possible population responses. Hence, lots of presentations were centered around ways to analyze high-dimensional data.

Selected neuroinformatics resources:

  • Brain Modeling Toolkit (BMTK), an open-source software package for creating and simulating large-scale brain models in Python. (Allen Institute for Brain Science)
  • MNI Open iEEG Atlas, an open access atlas of normal intracranial EEG (iEEG) of the human brain durig wakefulness (106 subjects, 1785 channels, registered to a common stereotaxic space). (Montreal Neurological Institute and Hospital, Grenoble-Alpes University Hospital, Centre hospitalier de l’Université de Montréal)
  • Drift correction for electrophysiology and two-photon calcium imaging, made available through their open-source packages Suite2P and KiloSort. (Marius Pachitariu et al., HHMI Janelia)
  • AJILE dataset, a dataset of annotated joints in long-term ECoG. (Bing Brunton et al., University of Washington)
  • pulse2percept, a Python based simulation framework for bionic vision. (Michael Beyeler et al., University of Washington)
  • seqNMF, a software package that implements regularized convolutional nonnegative matrix factorization (NMF). (Mackevicius, MIT)

More information can be found in the PowerPoint slides at the bottom of this page.

The full meeting program can be found here.

Workshops

The full workshop program can be found here.

Resources