Periodic Reporting for period 4 - PRIORS (Neural circuit dynamics underlying expectation and their impact on the variability of perceptual choices)
Período documentado: 2021-03-01 hasta 2022-02-28
In this project we investigate the mechanisms underlying the perception of the environment. In particular we have investigated the neural basis of how experience shapes our perception, or how expectations formed by our previous actions and their outcomes, are (1) built, (2) combined with sensory information and (3) processed to yield decisions. Expectations is an often overlooked topic in sensory systems neuroscience, in which the hierarchical circuitry starting from sensory receptors all the way up to the cortex has been finely characterized in various modalities. Current models of perceptual decision making assume that, during psychophysical tasks in which subjects need to discriminate ambiguous stimuli, a sensory representation formed in early cortical areas is integrated by associative areas in order to yield choices. The choice variability ubiquitous to all choices rising from stimuli at the perceptual threshold is thought to come from neuronal variability at the level of population of cortical sensory neurons. We hypothesize however, that a fraction of this variability might be caused by variability in the prior expectations formed by subjects based on previous decisions.
Overall we have been able to characterize the expectation biases developed by rats and humans during perceptual tasks with sequences of trials. Using inactivation experiments, we have characterized which are the brain areas involved in their computation and how these expectations are combined with sensory evidence in order to yield decisions. We have characterized this behavior using behavioral modeling and have leveraged on the behavioral models to analyze the electrophysiological signals recorded from the dorsal striatum and the prefrontal cortex while the animals did the task. Our findings will shed light onto expectations, an instrumental part of our perceptual experiences which may be altered in certain brain pathologies.
Overall we have been able to characterize the expectation biases developed by rats and humans during perceptual tasks with sequences of trials. Using inactivation experiments, we have characterized which are the brain areas involved in their computation and how these expectations are combined with sensory evidence in order to yield decisions. We have characterized this behavior using behavioral modeling and have leveraged on the behavioral models to analyze the electrophysiological signals recorded from the dorsal striatum and the prefrontal cortex while the animals did the task. Our findings will shed light onto expectations, an instrumental part of our perceptual experiences which may be altered in certain brain pathologies.
During the project we have published five main scientific articles:
1. Response outcomes gate the impact of expectations on perceptual decisions. A Hermoso-Mendizabal*, a Hyafil*, P. E. Rueda-Orozco, S. Jaramillo, D. Robbe, J. de la Rocha (*equal contribution). Nat Comm 11, 1057, 2020
2. Proactive and reactive accumulation-to-bound processes compete during perceptual decisions. Lluís Hernández-Navarro, A. Hermoso-Mendizabal, D. Duque, J. de la Rocha*, A. Hyafil* (* equal contribution). Nat Comm, 12(1), 1-15. 2021.
3. Isolating perceptual biases caused by trial history during auditory categorization. D. Duque and J. de la Rocha. BioRxiv 2022.01.17.476581; doi: 2nd revision in J Neurosci.
4. Flexible categorization in perceptual decision making. Genís Prat-Ortega, Klaus Wimmer, Alex Roxin* and J. de la Rocha* (*equal contribution). Nat Comm 12, 1283, 2021.
5. Pre-training RNNs in ecologically relevant tasks explains sub-optimal behavioral reset. M. Molano-Manzón, D. Duque, G.R. Yang and J. de la Rocha. BioRxiv 2021.05.15.444287. (submitted to eLife).
However, there are a number of subprojects do not have the form of an article yet but will be submitted as scientific articles in the coming 1-2 years. This is a list of those that have a closer estimated submission date:
6. Plan, initiate and update: a model for rats' choice trajectories during perceptual decision tasks. Jordi Pastor+, Alex Garcia Durán+, Lluís Hernández-Navarro, D. Duque, Lejla Bektic, M. Molano-Mazón*, A. Hyafil* and J. de la Rocha* (+equal contribution) (*equal senior contribution)(estimated submission March 2023; targeted journal Curr Biol).
7. Reinforcement learning of abstract rules involves the prefrontal cortex and the striatum. R. Marín+, C. Sindreu+, A. Hermoso-Mendizabal+, D. Duque, M. Molano-Mazón, L. Bektic, D. Robbe, A. Hyafil* and J. de la Rocha*. (+equal contribution) (*equal senior contribution)(estimated submission December 2022; targeted journal Nat Neurosci).
8. Idiosyncratic Attractor dynamics during fixed-duration categorization tasks in rats and humans. G. Prat-Ortega, E. Fernàndez, C. Pericas, K. Wimmer and J. de la Rocha. (estimated submission July 2023; targeted journal eLife).
9. Neural traces of decision vigor in the dorso medial striatum. Lluís Hernàndez-Navarro, A. Hermoso-Mendizabal, D. Duque, J. de la Rocha* and A. Hyafil*. (*equal senior contribution)(estimated submission December 2022; targeted journal Nat Comm).
10. Network dynamics underlying flexible gatting of context information. M. Molano-Mazón+, Y. Shao+, A. Hermoso-Mendizabal, L. Bektic, S. Ostojic* and J. de la Rocha*. (+equal contribution) (*equal senior contribution)(estimated submission date July 2023; targeted journal Nature Neuroscience).
In total, the work can be classified in three main branches: (i) animal experiments, (ii) human experiments and (iii) computational models. Each include different methodologies that we summarize in Table 1. Our scientific contributions have in almost all cases been driven by some type of experimental data we have acquired in the laboratory (e.g. behavioral data, electrophysiology, etc) and the development of some kind of computational model. In that respect, what makes our research original and impactful is this combination of experiments and modeling which can be clearly viewed in Table 1.
1. Response outcomes gate the impact of expectations on perceptual decisions. A Hermoso-Mendizabal*, a Hyafil*, P. E. Rueda-Orozco, S. Jaramillo, D. Robbe, J. de la Rocha (*equal contribution). Nat Comm 11, 1057, 2020
2. Proactive and reactive accumulation-to-bound processes compete during perceptual decisions. Lluís Hernández-Navarro, A. Hermoso-Mendizabal, D. Duque, J. de la Rocha*, A. Hyafil* (* equal contribution). Nat Comm, 12(1), 1-15. 2021.
3. Isolating perceptual biases caused by trial history during auditory categorization. D. Duque and J. de la Rocha. BioRxiv 2022.01.17.476581; doi: 2nd revision in J Neurosci.
4. Flexible categorization in perceptual decision making. Genís Prat-Ortega, Klaus Wimmer, Alex Roxin* and J. de la Rocha* (*equal contribution). Nat Comm 12, 1283, 2021.
5. Pre-training RNNs in ecologically relevant tasks explains sub-optimal behavioral reset. M. Molano-Manzón, D. Duque, G.R. Yang and J. de la Rocha. BioRxiv 2021.05.15.444287. (submitted to eLife).
However, there are a number of subprojects do not have the form of an article yet but will be submitted as scientific articles in the coming 1-2 years. This is a list of those that have a closer estimated submission date:
6. Plan, initiate and update: a model for rats' choice trajectories during perceptual decision tasks. Jordi Pastor+, Alex Garcia Durán+, Lluís Hernández-Navarro, D. Duque, Lejla Bektic, M. Molano-Mazón*, A. Hyafil* and J. de la Rocha* (+equal contribution) (*equal senior contribution)(estimated submission March 2023; targeted journal Curr Biol).
7. Reinforcement learning of abstract rules involves the prefrontal cortex and the striatum. R. Marín+, C. Sindreu+, A. Hermoso-Mendizabal+, D. Duque, M. Molano-Mazón, L. Bektic, D. Robbe, A. Hyafil* and J. de la Rocha*. (+equal contribution) (*equal senior contribution)(estimated submission December 2022; targeted journal Nat Neurosci).
8. Idiosyncratic Attractor dynamics during fixed-duration categorization tasks in rats and humans. G. Prat-Ortega, E. Fernàndez, C. Pericas, K. Wimmer and J. de la Rocha. (estimated submission July 2023; targeted journal eLife).
9. Neural traces of decision vigor in the dorso medial striatum. Lluís Hernàndez-Navarro, A. Hermoso-Mendizabal, D. Duque, J. de la Rocha* and A. Hyafil*. (*equal senior contribution)(estimated submission December 2022; targeted journal Nat Comm).
10. Network dynamics underlying flexible gatting of context information. M. Molano-Mazón+, Y. Shao+, A. Hermoso-Mendizabal, L. Bektic, S. Ostojic* and J. de la Rocha*. (+equal contribution) (*equal senior contribution)(estimated submission date July 2023; targeted journal Nature Neuroscience).
In total, the work can be classified in three main branches: (i) animal experiments, (ii) human experiments and (iii) computational models. Each include different methodologies that we summarize in Table 1. Our scientific contributions have in almost all cases been driven by some type of experimental data we have acquired in the laboratory (e.g. behavioral data, electrophysiology, etc) and the development of some kind of computational model. In that respect, what makes our research original and impactful is this combination of experiments and modeling which can be clearly viewed in Table 1.
1. We have modeled the history biases of rats during a 2AFC task much beyond the state of the art. We have been the first to separate the estimated probabilities animals may have about future events from the impact those predictions may have on future choices (i.e. in our case via choice biases).
2. We have exploited the analysis of the movement of the animals during the task using novel published methods (e.g. DeepLabCut). We have been able to extract much more information from the behavior of the animals than standard methods based on nose pokes or tracking the overall position of the animal. For instance, we are now able to identify trials in which animals made changes of mind (CoM) by characterizing the fine orienting movements of their snouts towards one side port or the other. We are currently training deep convolutional networks to try to identify choice biases in the body position when they start a new trial (before taking the decision).
3. Our electrophysiology data from dorso-medial striatum and frontal orienting field has put forward the role of the basal ganglia in decision making and particularly in mediating the predictions and choice biases computed based on previous trials and maintained in memory during several trials in order to condition future choices. We have found several different types of neuronal population which, to our knowledge, had not been described.
4. We have developed the Mouse Village, an open source behavioral platform in which animals are housed and trained in a self-paced manner, without any human intervention. In the MV groups of mice live freely in enriched arenas simulating their natural burrow systems, displaying complex social interactions while progressively learning cognitive tasks in an all-purpose touchscreen operant box connected to the system. We have recently submitted a ERC-PoC22 proposal to improve and distribute the pilot version of the MV developed under the ERC-CoG-2015-PRIORS.
2. We have exploited the analysis of the movement of the animals during the task using novel published methods (e.g. DeepLabCut). We have been able to extract much more information from the behavior of the animals than standard methods based on nose pokes or tracking the overall position of the animal. For instance, we are now able to identify trials in which animals made changes of mind (CoM) by characterizing the fine orienting movements of their snouts towards one side port or the other. We are currently training deep convolutional networks to try to identify choice biases in the body position when they start a new trial (before taking the decision).
3. Our electrophysiology data from dorso-medial striatum and frontal orienting field has put forward the role of the basal ganglia in decision making and particularly in mediating the predictions and choice biases computed based on previous trials and maintained in memory during several trials in order to condition future choices. We have found several different types of neuronal population which, to our knowledge, had not been described.
4. We have developed the Mouse Village, an open source behavioral platform in which animals are housed and trained in a self-paced manner, without any human intervention. In the MV groups of mice live freely in enriched arenas simulating their natural burrow systems, displaying complex social interactions while progressively learning cognitive tasks in an all-purpose touchscreen operant box connected to the system. We have recently submitted a ERC-PoC22 proposal to improve and distribute the pilot version of the MV developed under the ERC-CoG-2015-PRIORS.