Skip to main content

Coronavirus (COVID-19): Visitor Restrictions, Resources, and Updates

Explore URMC

FY 2017-18 Winner Conference reports

Computational and Systems Neuroscience Annual Meeting

  • Main meeting: 01--04 March 2018 in Denver, Colorado
  • Workshops: 05--06 March 2018 in Breckenridge, Colorado

Report Author

Marjena Popović, Postdoctoral fellow, DeAngelis lab
Center for Visual Science, University of Rochester

CoSyNe (Computational and Systems Neuroscience) is an annual meeting of scientists from the fields of computational and systems neuroscience, computer science, and machine learning. It was founded in 2004 by the first generation of computational neuroscientists: Carlos Brody (Cold Spring Harbor Research Laboratories at the time, now Princeton), Alexandre Pouget (University of Rochester at the time, now University of Geneva), Michael Shadlen (University of Washington at the time, now Columbia University), Tony Zador (Cold Spring Harbor Research Laboratories), Zach Mainen (Cold Spring Harbor Research Laboratories at the time, now Champalimaud Centre for the Unknown). It was envisioned as a forum that would bring together theoreticians and experimentalists, the first of its kind at the time. Until this day CoSyNe remains one of the most important meetings for computational and systems neuroscience.

The conference is divided into two sections: the main conference (03/01 - 03/04/18) and workshops (03/04 - 03/06/18). This year, for the first time, CoSyNe opened with a tutorial session on statistical modeling techniques for neural signals titled “Statistical models for neural data: from GLMs to latent variables”, organized by Memming Park (Stony Brook University) and lead by Jonathan Pillow (Princeton University). The organizers have received overwhelmingly positive feedback about the tutorial session and plan to organize similar events at future instances of CoSyNe. Slides for the tutorial have been made publicly available and can be found on Jonathan Pillow’s website.

Main meeting is single track and features both invited and submitted talks. Somewhat unusually, submissions are evaluated based on a 300-word summary and additional detail that may include figures and equations taking up to 2 pages. This is necessary because of high rejection rates at CoSyNe: this year there were 857 registrations, 709 abstracts were submitted, and 396 accepted - 55.8%. The high rejection rate leads to an exceptionally high quality of talks as well as abstracts, one that is in my opinion unparalleled by any other conference in neuroscience. As a result, most conference attendees attend all the talks. CoSyNe is known for having a very demanding schedule: talks start at 8:30am and go on until 5:30pm, with a few short coffee breaks and a 2 hour lunch break in between; the posters start at 8pm and officially end at 11pm, though it is not rare to see people engaged in discussions in the poster hall past midnight. CoSyNe is also known for having a presence on twitter and for being covered on blogs of several attendees.

This year CoSyNe organized the first lunch/workshop for Equality and Diversity in Science on March 3rd. It opened with an introduction by the organizers Maria Geffen (University of Pennsylvania) and Stephanie Palmer (University of Chicago), followed by presentations with practical advice on how to overcome and respond to bias in science and focused round-table discussions and networking. The presenters were Ann Graybiel (MIT), Marlene Cohen (University of Pittsburg), Jess Cardin (Yale School of Medicine) and Linda Wilbrecht (UC Berkeley).

The presentations at the main meeting were dominated by a dynamical systems view of the brain. Individual neuronal responses populate space with a very high number of dimensions. As a result, neural data often very sparsely populates that space which leads to problems in generating predictions from computational models of neural responses and problems in model selection due to overfitting. One of the focuses of this year’s main meeting was dimensionality reduction techniques – specifically ways of finding relevant lower dimensional manifolds[1] in which the neural responses have the highest variance. Byron Yu (Carnegie Melon University) gave an interesting talk that used the concept of low dimensional manifolds to explore how animals learn to use brain computer interfaces on short time scales.

The absolute highlight of CoSyNe this year was the Gatsby lecture by Iain D. Couzin (Max Planck Institute at University of Konstanz) titled “Collective sensing and decision-making in animal groups: From fish schools to primate societies”. Ian presented a large body of work he has done during his research career, showing how he went from building theoretical models of swarm behavior using schools of fish, to tracking troops of baboons in the wild using the global positioning system. He showed that group decisions in schools of fish don’t require signaling, individual recognition or knowledge of the preferences of others, and that global decisions can be made using only local decision rules, i.e. the group can have awareness even though the individuals do not. Interestingly, group decisions are not swayed more by “popular” individuals, but by individuals who themselves have a low degree of connectivity and/or are connected to individuals with a high degree of connectivity. Furthermore, he showed that uninformed individuals contribute to the decision-making process by preventing an informative minority from overwhelming the decision making, effectively ensuring that decisions are made by the majority. Because animals that are closer together tend to have a similar “opinion”, an increase in-group size doesn’t lead to making better decisions. Though interesting on its own, this work has an important implication for how we try to understand neural data. This work shows that to understand the behavior of an individual, one has to (first) understand the behavior of the group.

Another talk that garnered a lot of attention was given by Carina Curto (Pennsylvania State University) and titled “Emergent dynamics from network connectivity: a minimal model”. Carina talked about how interesting dynamics emerge from networks that are simple enough to remain mathematically tractable. Though this was a talk about applied mathematics and not expressly geared toward an audience of neuroscientists it was very well received and generated a lot of discussion among participants.

As is now expected from CoSyNe, this year we saw presentations of a wealth of open source neural data processing toolkits and open access datasets and atlases. The neuroinformatics working group at the University of Washington (UW) eScience Institute and the University of Washington Institute for Neuroengineering (UWIN) compiled a list of these resources and published them on their blog:

  • 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 during 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)

[1] topological space that locally resembles Euclidean space near each point

The International Association of Yoga Therapists Symposium of Yoga Research  

  • October 16-18th, 2017
  • Kripalu Center for Yoga & Health Center, Stockbridge MA

Report Author 

Po-Ju Lin, Ph.D., Postdoctoral Fellow
PEAK Human Performance Clinical Research Core Laboratory, University of Rochester

The international Association of Yoga Therapists (IAYT) was founded in 1989 and is dedicated to establishing yoga as a recognized and respected therapy.  IAYT supports research and education in yoga and serves as a professional organization for yoga instructors and therapists worldwide.  The Symposium on Yoga Research (SYR) is the foremost academic yoga research conference in the West and is one of the yoga conferences held by IAYT. SYR showcases the updated yoga research and offers professional opportunities to interact with scientists, experts, and professionals in the field of yoga research.  

I am grateful to be awarded with 2017-18 Travel and Professional Development Award from the Postdoctoral Association at the University of Rochester to support me attending 2017 SYR and presenting my research work.  My research entitled “The mediational relationship between sleep and cancer-related fatigue stemming from yoga: A URCC NCORP RCT in 321 cancer patients” examined the mediational effect of improved sleep quality stemming from yoga on the changes in cancer-related fatigue in 321 cancer survivors.  My findings indicate that improvements in sleep quality, particularly via reducing daytime napping, contributed to 22-37% of the improvements in cancer-related fatigue resulting from a four-week yoga program in cancer survivors. This study is among the very first studies to examine whether the effect of yoga on reducing cancer-related fatigue is mediated by improvements in sleep quality and it was recognized with the research award by the IAYT.

By attending the IAYT SYR conference, I gained exposure for myself and my research, and was recognized by others in the field.  I learned a lot from presentations delivered by keynote speakers and other well-known yoga researchers.  Dr. Lorenzo Cohen, professor and director of integrative medicine program at the University of Texas MD Anderson Cancer Center, presented a comprehensive approach to transform your life and health.  He summarized the preventable causes of cancer and how mind-body practices (e.g. yoga, qigong, tai chi) can help manage cancer-related side effects and toxicities.  His research also showed that yoga improved sleep quality, sleep efficiency, and daytime dysfunction (daytime napping) in breast cancer patients during and after chemotherapy.  Dr. Nicole Culos-Reed from the University of Calgary pointed out limitations from current yoga literature such as lack of large scale randomized clinical trials, various styles, doses, and protocols of the yoga interventions, recruitment and retention issues, and the generalizability to patients with different types and stages of cancer and discussed the importance of the knowledge translation between what we know about yoga in cancer care and what is being offered to cancer patients and survivors clinically.  Dr. Stephanie Sohl from Wake Forest Baptist Medical Center raised the issue of the large gap between research and practice.  She focused on the feasibility of implemented yoga intervention for cancer patients during chemotherapy and around the time of surgery.  All of the studies presented in SYR were fascinating and have provided me ideas, skills, and research direction for future studies.  Most importantly, I had the opportunity to personally interact with these well-known yoga researchers to ask questions and discuss the potential of my research ideas and study design and delivery.

The University of Rochester Postdoctoral Association Travel and Professional Development Award provides me the financial support to present my yoga research in the foremost academic yoga research conference and interact with other researchers in the field.  This opportunity is extremely helpful for young investigators like me to exposure their research work, build the reputation, network and seek possible cooperation, and help academic career development. I truly appreciate being awarded with the Travel and Professional Development Award provided by the Postdoctoral Association and have received the most benefits for my professional development.  I strongly recommend other postdoctoral fellows at University of Rochester take this great opportunity to apply for the Travel and Professional Development Award.