Eftychios Pnevmatikakis joined the Simons Foundation after working as a postdoctoral research scientist in the department of statistics and the Center for Theoretical Neuroscience at Columbia University. During his postdoctoral research, Pnevmatikakis created efficient machine-learning and optimization algorithms for statistical analysis of large-scale electrophysiology and optical-physiology data. As part of CCB’s neuroscience group, Pnevmatikakis develops methods and software for analysis of large sets of neuroscience data, as well as theories of neural computation. He holds a Ph.D. in electrical engineering from Columbia University.
My research has primarily focused on the development of novel tools in the fields of statistics, machine learning, signal processing and optimization, both from a theoretical perspective and from a practical standpoint, with applications related to neural data analysis and neural computation. I am particularly interested in the development of efficient algorithms for regularization and analysis of high-dimensional datasets, a pressing need due to the ongoing explosion of data from experimental neuroscience. I’m currently working on developing novel methods and software for analyzing calcium imaging data.
Calcium imaging methods have revolutionized data acquisition in neuroscience; we can now record from large neural populations to study the structure and function of neural circuits, or from multiple specific locations on a dendritic tree to examine the detailed computations performed at a subcellular level. However, this new data source has its drawbacks; the fast neural activity is observed indirectly by a slow and noisy calcium signal, obtained typically at a low temporal rate. A basic goal of the data analyst is to deconvolve the underlying activity from the observed signal, and to identify and separate out the different sources. In my research I develop methods that address these problems in a computationally tractable way and collaborate with multiple experimental labs to address specific scientific questions that arise from these methods.
Ph.D., Electrical Engineering (2010)
Columbia University, New York, NY
5-Year Diploma, Electrical and Computer Engineering (2004)
National Technical University of Athens, Athens, Greece
Some of the software packages I’ve developed can be found on my public github page.
A complete list of publications can be found on my Google Scholar page.
Pnevmatikakis EA, Soudry D, Gao Y, Machado TA, … & Paninski L. Simultaneous denoising, deconvolution, and demixing of calcium imaging data. Neuron. 2016;89(2):285-299 http://dx.doi.org/10.1016/j.neuron.2015.11.037.
Yang W Miller JE, Carrillo-Reid L, Pnevmatikakis EA, Paninski L, Yuste R, Peterka DS. Simultaneous Multi-plane Imaging of Neural Circuits. Neuron. 2016;89(2):269-284 http://dx.doi.org/10.1016/j.
Machado TA, Pnevmatikakis EA, Paninski L, Jessell TM, Miri A. Primacy of flexor locomotor pattern revealed by ancestral reversion of motor neuron identity. Cell. 2015;162(2):338-350. doi:10.1016/j.cell.2015.06.036.
Ramirez A, Pnevmatikakis EA, Merel J, Paninski L, Miller KD, Bruno RM. Spatiotemporal receptive fields of barrel cortex revealed by reverse correlation of synaptic input. Nat Neurosci. 2014;17(6):866-875. doi:10.1038/nn.3720.
Pnevmatikakis EA, Gao Y, Soudry D, et al. A structured matrix factorization framework for large scale calcium imaging data analysis. arXiv preprint 2014; arXiv:1409.2903. http://arxiv.org/abs/1409.2903.
Pnevmatikakis EA, Rad KR, Huggins J, Paninski L. Fast Kalman filtering and forward-backward smoothing via a low-rank perturbative approach. J Comput Graph Stat. 2014;23(2):316-339. doi:10.1080/10618600.2012.760461.
Pnevmatikakis EA, Paninski L. Sparse nonnegative deconvolution for compressive calcium imaging: Algorithms and phase transitions. In: Burges CJC, Bottou L, Welling M, Ghahramani Z, Weinberger KQ, eds. Advances in Neural Information Processing Systems (NIPS) 2013. Neural Information Processing Systems Foundation; 2013:1250-1258. http://papers.nips.cc/paper/4996-sparse-nonnegative-deconvolution-for-compressive-calcium-imaging-algorithms.
Pfau D, Pnevmatikakis EA, Paninski L. Robust learning of low-dimensional dynamics from large neural ensembles. In: Burges CJC, Bottou L, Welling M, Ghahramani Z, Weinberger KQ, eds. Advances in Neural Information Processing Systems (NIPS) 2013. Neural Information Processing Systems Foundation; 2013:2391-2399. http://papers.nips.cc/paper/4995-robust-learning-of-low-dimensional-dynamics-from-large-neural-ensembles.
Pnevmatikakis E, Merel J, Pakman A, Paninski L. Bayesian spike inference from calcium imaging data. In: Matthews MB, ed. Conference Record of the 47th Asilomar Conference on Signals, Systems and Computers. Institute of Electrical and Electronics Engineers; 2013;349-353. doi:10.1109/ACSSC.2013.6810293.
Pnevmatikakis EA, Kelleher K, Chen R, Saggau P, Josic K, Paninski L. Fast spatiotemporal smoothing of calcium measurements in dendritic trees. PLoS Comput Biol. 2012;8(6):e1002569. doi:10.1371/journal.pcbi.1002569.
Pnevmatikakis EA, Paninski L. Fast interior-point inference in high-dimensional sparse, penalized state-space models. Paper presented at: Fifteenth International Conference on Artificial Intelligence and Statistics (AISTATS); April 21-23, 2012; La Palma, Canary Islands. http://machinelearning.wustl.edu/mlpapers/paper_files/AISTATS2012_PnevmatikakisP12.pdf.
*Lazar A, *Pnevmatikakis E. Video time encoding machines. IEEE Trans Neural Netw. 2011;22(3):461-473. doi:10.1109/TNN.2010.2103323.
*Lazar AA, *Pnevmatikakis EA, *Zhou Y. Encoding natural scenes with neural circuits with random thresholds. Vision Res. 2010;50(22):2200-2212. doi:10.1016/j.visres.2010.03.015.
*Lazar AA, *Pnevmatikakis EA. Faithful representation of stimuli with a population of integrate-and-fire neurons. Neural Comput. 2008;20(11):2715-2744. doi:10.1162/neco.2008.06-07-559.
Pnevmatikakis EA, Maragos P. An inpainting system for automatic image structure-texture restoration with text removal. In: 2008 IEEE International Conference on Image Processing: ICIP 2008 Proceedings. Institute of Electrical and Electronics Engineers; 2008;2616-2619. doi:10.1109/ICIP.2008.4712330.
*indicates alphabetical listing of the authors’ names