My background in Physics and Mathematics and my training in Computational Neuroscience during my PhD provide my research on the neural circuit bases of behavior with the quantitative and mathematical substrate that modern neuroscience is demanding nowadays. Moreover, a solid training in electrophysiology during my postdoctoral years in the Center for Neural Science (NYU, 2004-08) and the Center for Molecular and Behavioral Neuroscience (Rutgers Univ., 2008-10) and the experience gained in animal behavior in my own laboratory (IDIBAPS, 2010-16) give me the skills to carry out a multi-lateral approach to characterize the dynamics of neural circuits during decision making.

Over the last eight years I have been using multi-electrode arrays to perform large scale recordings from rat cortex. I have combined these recordings with advanced statistical analysis and computational modeling to reveal how spontaneous activity changes with brain state and it can impact the representation of sensory information [1-2]. In the last four years I have focused on studying decision making experiments in rats combined with chronic population recordings and pharmacological inactivation of striatial circuits and circuits in the pre-frontal cortex [4]. Along these lines, I have characterized the relationship between decision-making behavior and the dynamics of cortical circuits [5]. For this purpose I use techniques of advanced statistical learning to finely quantify the dynamics of behavior. Combined with neural physiology, these statistical models are an essential tool to find the neural correlates of the different aspects of behavior (e.g. memory guided choice biases [6]). Moreover, I am currently performing similar psychophysical tasks in human subjects and will soon start a collaboration to combine these experiments with MEG recordings (collaboration with Prof. Tobias Donner, Univ. Hamburg, Germany). More recently we have started characterizing unveiling the neural circuit basis of long-term and working memory tasks in mice while recording the activity of large ensembles of neurons using one-photon Calcium imaging.

Finally, with support from a European Research Council Consolidator grant (ERC-CoG) that just started (Sept. 2016), my lab is setting up a number of cutting-edge techniques in the study of the neural basis of expectation biases in sensory discrimination tasks in rats and mice. This techniques include automatic high throughput mouse and rat training (i.e. computer-controlled arrays of multiple behavioral boxes, e.g. 10-20), optogenetic and pharmacologic inactivation of cortical brain circuits and pathways, wide field of view (5 x 5 mm) Calcium imaging in head fixed behaving mice and computational modeling of cortical circuits.

[1] Renart*, de la Rocha* et al. Science 327: 587-590, (2010)

[2] Mochol PNAS: 112.11 (2015): 3529-3534

[3] D. Jercog et al. UP/DOWN cortical dynamics reflect stochastic transitions in a bi-stable balanced network. BioRxiv 083626 (2016). (submitted to peer-review).

[4] Hermoso-Mendizabal et al. The impact of prior expectations in an auditory discrimination task. No. 651.07. Washington, DC: SfN 2014.

[5] Wimmer K. et al. Nat Comm, 6. (2015)

[6] Hayfil A et al. Modeling history dependent perceptual biases in rodents. ICANN'19 Proceed. 26th Int. Conf. Artificial Neural Networks.