I am using machine learning and data analytics for large-scale neuroscience, working on identifing cell types, circuits and computations in the healthy and diseased visual system. In 2016, I will set up my own group funded by the Bernstein Award. I collaborate closely with Matthias Bethge and Thomas Euler at the University of Tübingen and Andreas Tolias at Baylor College of Medicine. See my Research page for my current projects and interests.
Our new paper has been published today in Science! In this work lead by Xiaolong Jiang in Andreas Tolias’ lab at Baylor College of Medicine in Houston, we classified cortical interneurons in adult mice and mapped their connectivity diagram. Interestingly, three simple rules go a long way in explaining the structure of the connectivity matrix. For me, this is a nice piece of heroic wet-lab work combined with the right level of machine learning based analysis.
I just attended the European Retina Meeting in Brighton, which, surprisingly boasted 5 days of sunshine and blue skies (see right).
We had organized a sattelite meeting on retinal ganglion cell classification on the first day. Sebastian Seung presented his latest advances on morphological classification of RGCs (his museum can be found here). EJ Chichilnisky showed his advances on finding rare cell types in the macaque retina, before I presented our work on functional RGC classification. Then Josh Sanes introduced new techniques for obtained single cell transcriptomes and a proof of principle for retina bipolar cell type classification. Estuardo Robles shared his data on the role of RGC projections in defining cell types before Andy Huberman talked about possibilities to link cell types across multiple modalities. It was an exciting, discussion-rich day with lots of new information.
The main meeting was also very intense, with lots of great talks. In particular, Markus Meister presented convincing data on a new fourth alpha-like RGC type in the mouse retina, in good agreement with our data. The dominating topic was direction selectivity, with many interesting contributions.
My proposal “Cell type specific neural computations” was selected for this years Bernstein award. I am honored, grateful and looking forward to 5 years of exciting research opportunities.
The workshop on retinal computations Tom Baden and I organized took place this morning at BCCN2015 in Heidelberg. We had 5 great talks, about 50 participants, and quite lively discussions, making the workshop a great success! Martin Greschner shared his latest results on ganglion cell types in the guinnea pig retina. Karl Farrow talked about recent result related to the retinal computations at different light levels. Julijana Gjorgjieva shared her computational insights into the computational purpose of pathway splitting, before Katrin Franke showed our results on ganglion cell types in the mouse. Finally, Laura Busse presented her advances studying the effect of cortical feedback on LGN activity.
We have been organizing a sattelite symposium on retinal ganglion cell classification a day before the European Retina Meeting. It aims to bring together people with different approaches (anatomy, genetics, physiology, connectomics) to this topic and discuss the latest advances. I think we have an exciting program including talks by Sebastian Seung, EJ Chichilnisky, Joshua Sanes, Andy Huberman and Estuardo Robles. I will present the latests results from our endeavours into functionally classifying retinal ganglion cells. Space is pretty much filled up, but here
is the program.
Together with Lucas Theis, a PhD student in our lab, I developed a new method to infer spike rates from calcium signals using supervised learning in flexible probabilistic models. In addition, together with our experimental collaborators we collected a large dataset of simultaneously recorded electrophysiology and calcium imaging data that allowed us to benchmark our algorithm against many available methods.
We show that our algorithm performs better than previoulsy published methods (preprint) and make the code available on github. The package includes a model trained on all available data, which should also perform well without training data. Let us know when you use the code in your research!
Tom Baden and I are organizing a workshop at the Bernstein Conference 2015 on ‘What the eye tells the brain and why’. It will take place on the 14th of September, between 9 and 12.30 am as part of the Sattelite Workshop day. We will have Martin Greschner, Karl Farrow, Julijana Gjorgjieva, Katrin Franke and Laura Busse presenting the latest on coding in the retina and beyond! We are looking forward to an exciting day, especially since there are two related afternoon workshops on ‘Dynamic retinal coding’ and ‘From retina to robots’.
Christian Behrens, a new PhD student, has joined our lab to work with me on data analysis and modeling in the retina. He studied theoretical physics in Munich and Heidelberg before coming to Tübingen. He will join the team working on the DFG grant we have been recently awarded, collaborating with Tom Baden, Katrin Franke and Thomas Euler.
We have recently started to implement PyCircStat, a python package for statistical analysis of circular or angular data. It builds on the matlab toolbox for circular statistics I have developed a few years ago. The python package is still in the early stages of development, but a number of features are already working:
You are welcome to check out the toolbox and try it – let us know if you find bugs, or have feature requests. Thanks!