Behavior arises from the coordinated activity across anatomically and functionally distinct brain regions. Modern experimental tools allow unprecedented access to large neural populations spanning many interacting regions brain-wide. Yet, understanding such large-scale datasets necessitates robust, scalable computational models to extract meaningful features of inter-region communication and principled theories to interpret those features. Here, we introduce current-based decomposition (CURBD), an approach for inferring brain-wide interactions using data-constrained recurrent neural network models that autonomously produce dynamics consistent with experimentally obtained neural data. CURBD leverages the functional interactions inferred from such models to reveal directional currents between multiple brain regions simultaneously. We first show that CURBD accurately isolates inter-region currents in simulated, ground-truth networks with known connectivity and dynamics. We then apply CURBD to multi-region neural recordings obtained from many species-larval zebrafish, mice, macaques, and humans-to demonstrate the widespread applicability of CURBD in untangling brain-wide interactions and inter-area communication principles underlying behavior.

Introduction: Hepatogenous diabetes (HD) is a disorder of glucose metabolism (DGM) that develops as a complication of advanced chronic liver disease (ACLD) with an estimated prevalence of 20–70%. It appears to be associated with a larger number of decompensations, but its impact on the natural history of the disease is unclear. The treatment of DGM is hampered by the fact that some therapeutic agents are associated with a risk of complications in ACLD. The aim of this work was to study DGM in a population of patients with ACLD: prevalence, liver disease decompensation episodes, mortality analysis and study of the impact of antidiabetic therapies in patients with ACLD who developed DGM. Materials and Methods: A cohort of consecutive patients with ACLD without previous DGM, who attended a Hepatology clinic in the period of January to June 2015 was selected. Follow-up was carried out for 5 years. Data on age, gender, date of diagnosis and etiology of ACLD, Child-Pugh and MELD-Na classifications at enrollment, development of DGM, and antidiabetic therapy were collected. Logistic regression models for hospitalizations due to decompensated ACLD, ascites, hepatic encephalopathy (HE), upper gastrointestinal bleeding (UGB), hepatocellular carcinoma (HCC), portal vein thrombosis (PVT), infectious complications, acute-on-chronic liver failure (ACLF), and death were built. A survival analysis for patients with and without DGM was also performed. Treatment effectiveness for patients with DGM was assessed. Results: Initially, 221 patients were included, 154 (69.7%) of whom developed DGM after the diagnosis of ACLD. DGM patients presented a significantly higher number of hospitalizations. Odds ratio (OR) for death was not significantly related with DGM. At 5 years of follow-up, 68.9% of patients with DGM were alive, against 81.8% without DGM (p = 0.087). From the 154 patients who were diagnosed with DGM, 42.9% were not receiving pharmacological treatment for DGM. Treated patients were prescribed with either biguanides (34.8%), a SGLT2 inhibitor (8.6%), or insulin (7.7%). Only 1 patient was treated with a GLP-1 analogue. A tendency of OR favoring treatment was observed for biguanides and SGLT2 inhibitors in all outcomes except ascites. In the univariable analysis, the use of biguanides was associated with lower risk of death (OR: 0.84 [95% CI: 0.73–0.96]) and HE (OR: 0.85 [95% CI: 0.73–0.98]). Conclusion: DGM occurs with high prevalence in patients with ACLD and it seems to be related to more hospitalizations, which highlights the importance of its early identification and appropriate therapeutic approach. In the absence of contraindications, biguanides should be considered for treatment of patients with ACLD and DGM as they appear to be associated with a tendency to better outcomes and may present some advantage in terms of survival.

Characterizing animal behavior requires methods to distill 3D movements from video data. Though keypoint tracking has emerged as a widely used solution to this problem, it only provides a limited view of pose, reducing the body of an animal to a sparse set of experimenter-defined points. To more completely capture 3D pose, recent studies have fit 3D mesh models to subjects in image and video data. However, despite the importance of mice as a model organism in neuroscience research, these methods have not been applied to the 3D reconstruction of mouse behavior. Here, we present ArMo, an articulated mesh model of the laboratory mouse, and demonstrate its application to multi-camera recordings of head-fixed mice running on a spherical treadmill. Using an end-to-end gradient based optimization procedure, we fit the shape and pose of a dense 3D mouse model to data-derived keypoint and point cloud observations. The resulting reconstructions capture the shape of the animal’s surface while compactly summarizing its movements as a time series of 3D skeletal joint angles. ArMo therefore provides a novel alternative to the sparse representations of pose more commonly used in neuroscience research.

The direct and indirect pathways of the basal ganglia are classically thought to promote and suppress action, respectively(1). However, the observed co-activation of striatal direct and indirect medium spiny neurons(2) (dMSNs and iMSNs, respectively) has challenged this view. Here we study these circuits in mice performing an interval categorization task that requires a series of self-initiated and cued actions and, critically, a sustained period of dynamic action suppression. Although movement produced the co-activation of iMSNs and dMSNs in the sensorimotor, dorsolateral striatum (DLS), fibre photometry and photo-identified electrophysiological recordings revealed signatures of functional opponency between the two pathways during action suppression. Notably, optogenetic inhibition showed that DLS circuits were largely engaged to suppress-and not promote-action. Specifically, iMSNs on a given hemisphere were dynamically engaged to suppress tempting contralateral action. To understand how such regionally specific circuit function arose, we constructed a computational reinforcement learning model that reproduced key features of behaviour, neural activity and optogenetic inhibition. The model predicted that parallel striatal circuits outside the DLS learned the action-promoting functions, generating the temptation to act. Consistent with this, optogenetic inhibition experiments revealed that dMSNs in the associative, dorsomedial striatum, in contrast to those in the DLS, promote contralateral actions. These data highlight how opponent interactions between multiple circuit- and region-specific basal ganglia processes can lead to behavioural control, and establish a critical role for the sensorimotor indirect pathway in the proactive suppression of tempting actions.

Neural activity in the mammalian cortex has been studied extensively during decision tasks, and recent work aims to identify under what conditions cortex is actually necessary for these tasks. We discovered that mice with distinct cognitive experiences, beyond sensory and motor learning, use different cortical areas and neural activity patterns to solve the same navigation decision task, revealing past learning as a critical determinant of whether cortex is necessary for goal-directed navigation. We used optogenetics and calcium imaging to study the necessity and neural activity of multiple cortical areas in mice with different training histories. Posterior parietal cortex and retrosplenial cortex were mostly dispensable for accurate performance of a simple navigation task. In contrast, these areas were essential for the same simple task when mice were previously trained on complex tasks with delay periods or association switches. Multiarea calcium imaging showed that, in mice with complex-task experience, single-neuron activity had higher selectivity and neuron-neuron correlations were weaker, leading to codes with higher task information. Therefore, past experience is a key factor in determining whether cortical areas have a causal role in goal-directed navigation.

Reward expectations based on internal knowledge of the external environment are a core component of adaptive behavior. However, internal knowledge may be inaccurate or incomplete due to errors in sensory measurements. Some features of the environment may also be encoded inaccurately to minimize representational costs associated with their processing. In this study, we investigated how reward expectations are affected by features of internal representations by studying behavior and dopaminergic activity while mice make time-based decisions. We show that several possible representations allow a reinforcement learning agent to model animals’ overall performance during the task. However, only a small subset of highly compressed representations simultaneously reproduced the co-variability in animals’ choice behavior and dopaminergic activity. Strikingly, these representations predict an unusual distribution of response times that closely match animals’ behavior. These results inform how constraints of representational efficiency may be expressed in encoding representations of dynamic cognitive variables used for reward-based computations.

Behavior arises from the coordinated activity of numerous anatomically and functionally distinct brain regions. Modern experimental tools allow unprecedented access to large neural populations spanning many interacting regions brain-wide. Yet, understanding such large-scale datasets necessitates both scalable computational models to extract meaningful features of inter-region communication and principled theories to interpret those features. Here, we introduce Current-Based Decomposition (CURBD), an approach for inferring brain-wide interactions using data-constrained recurrent neural network models that directly reproduce experimentally-obtained neural data. CURBD leverages the functional interactions inferred by such models to reveal directional currents between multiple brain regions. We first show that CURBD accurately isolates inter-region currents in simulated networks with known dynamics. We then apply CURBD to multi-region neural recordings obtained from mice during running, macaques during Pavlovian conditioning, and humans during memory retrieval to demonstrate the widespread applicability of CURBD to untangle brain-wide interactions underlying behavior from a variety of neural datasets. ### Competing Interest Statement The authors have declared no competing interest.

Reward expectations based on internal knowledge of the external environment are a core component of adaptive behavior. However, internal knowledge may be inaccurate or incomplete due to errors in sensory measurements. Some features of the environment may also be encoded inaccurately to minimise representational costs associated with their processing. We investigate how reward expectations are affected by differences in internal representations by studying rodents’ behaviour and dopaminergic activity while they make time based decisions. Several possible representations allow a reinforcement learning agent to model animals’ choices during the task. However, only a small subset of highly compressed representations simultaneously reproduce, both, animals’ behaviour and dopaminergic activity. Strikingly, these representations predict an unusual distribution of response times that closely matches animals’ behaviour. These results can inform how constraints of representational efficiency may be expressed in encoding representations of dynamic cognitive variables used for reward based computations.

Imbalance between action suppression and production characterizes several basal ganglia (BG) disorders. Relatedly, the direct and indirect pathways of the BG are hypothesized to promote and suppress actions, respectively. Yet striatal direct (dMSNs) and indirect (iMSNs) medium spiny neurons are coactive around movement, apparently contradicting direct-indirect functional opponency. In the dorsolateral striatum of mice, we observed coactivation around movements, but elevated and diminished activity of iMSNs and dMSNs, respectively, during action suppression. Furthermore, relative activity of the two hemispheres evolved in opposite directions in the two pathways as the need to suppress movements to either side of the body developed over time. Lastly, optogenetic inhibition experiments revealed the necessity of iMSNs but not dMSNs for the proactive suppression of specific actions, and dMSNs but not iMSNs for generalized action vigor. These data demonstrate distinct yet still broadly opponent roles for the direct and indirect pathways in behavioral control. ### Competing Interest Statement The authors have declared no competing interest.

Research in neuroscience increasingly relies on the mouse, a mammalian species that affords unparalleled genetic tractability and brain atlases. Here, we introduce high-yield methods for probing mouse visual decisions. Mice are head-fixed, facilitating repeatable visual stimulation, eye tracking, and brain access. They turn a steering wheel to make two alternative choices, forced or unforced. Learning is rapid thanks to intuitive coupling of stimuli to wheel position. The mouse decisions deliver high-quality psychometric curves for detection and discrimination and conform to the predictions of a simple probabilistic observer model. The task is readily paired with two-photon imaging of cortical activity. Optogenetic inactivation reveals that the task requires mice to use their visual cortex. Mice are motivated to perform the task by fluid reward or optogenetic stimulation of dopamine neurons. This stimulation elicits a larger number of trials and faster learning. These methods provide a platform to accurately probe mouse vision and its neural basis.

Our sense of time is far from constant. For instance, time flies when we are having fun, and it slows to a trickle when we are bored. Midbrain dopamine neurons have been implicated in variable time estimation. However, a direct link between signals carried by dopamine neurons and temporal judgments is lacking. We measured and manipulated the activity of dopamine neurons as mice judged the duration of time intervals. We found that pharmacogenetic suppression of dopamine neurons decreased behavioral sensitivity to time and that dopamine neurons encoded information about trial-to-trial variability in time estimates. Last, we found that transient activation or inhibition of dopamine neurons was sufficient to slow down or speed up time estimation, respectively. Dopamine neuron activity thus reflects and can directly control the judgment of time.

Full text Figures and data Side by side Abstract eLife digest Introduction Results Discussion Materials and methods References Decision letter Author response Article and author information Metrics Abstract The striatum is an input structure of the basal ganglia implicated in several time-dependent functions including reinforcement learning, decision making, and interval timing. To determine whether striatal ensembles drive subjects' judgments of duration, we manipulated and recorded from striatal neurons in rats performing a duration categorization psychophysical task. We found that the dynamics of striatal neurons predicted duration judgments, and that simultaneously recorded ensembles could judge duration as well as the animal. Furthermore, striatal neurons were necessary for duration judgments, as muscimol infusions produced a specific impairment in animals' duration sensitivity. Lastly, we show that time as encoded by striatal populations ran faster or slower when rats judged a duration as longer or shorter, respectively. These results demonstrate that the speed with which striatal population state changes supports the fundamental ability of animals to judge the passage of time. https://doi.org/10.7554/eLife.11386.001 eLife digest You know someone is a good cook from their rice - grains must be well cooked, but not to the point of being mushy. Despite consistently using the same pot and stove, we, however, will sometimes overcook it. It is as if our inner sense of time itself is variable. What is it about the brain that explains this variability in time estimation and indeed our ability to estimate time in the first place? One issue the brain must confront in order to estimate time is that individual brain cells typically fire in bursts that last for tens of milliseconds. So how does the brain use this short-lived activity to track minutes and hours? One possibility is that individual neurons in a given brain region are programmed to fire at different points in time. The overall firing pattern of a group of neurons will therefore change in a predictable way as time passes. Gouvêa, Monteiro et al. found such predictably changing patterns of activity in the striatum of rats trained to estimate and categorize the duration of time intervals as longer or shorter than 1.5 seconds. Interestingly, when rats mistakenly categorized a short interval as a long one, population activity had travelled farther down its path than it would normally (and vice-versa for long intervals incorrectly categorized as short), suggesting that variability in subjective estimates of the passage of time might arise from variability in the speed of a changing pattern of activity across groups of neurons. As further evidence for the involvement of the striatum, inactivating the structure impaired the rats’ ability to correctly classify even the longest and shortest interval durations. The next challenge is to determine exactly how the striatum generates these time-keeping signals, at which stage variability originates, and how the brain regions that the striatum signals to use them to control an animal’s behavior. https://doi.org/10.7554/eLife.11386.002 Introduction Time, like space, is a fundamental dimension of the environment, yet how it is processed in the brain is poorly understood. A number of recent studies have identified dynamics that allow for robust representation of time by populations of neurons in multiple areas including the hippocampus (Pastalkova et al., 2008; MacDonald et al., 2011), prefrontal (Kim et al., 2013; Xu et al., 2014), parietal (Leon and Shadlen, 2003; Janssen and Shadlen, 2005) and motor (Lebedev et al., 2008) cortices, cerebellum (Mauk and Buonomano, 2004), and the striatum (Matell et al., 2003; Jin et al., 2009; Adler, 2012; Mello et al., 2015). However, any dynamics that result in a continuously-evolving and non-repeating population state can be used to encode time (Buonomano, 2014), and it is not known whether such temporal representations would inform subjects’ judgments of duration or merely covary with elapsing time. The striatum, a brain structure known to be involved in reinforcement learning and decision making (Lau and Glimcher, 2008; Samejima et al., 2005; Lee et al., 2015), has been implicated in interval timing by several lines of evidence (Hinton and Meck, 2004; Harrington et al., 2009; Wencil et al., 2010; Malapani, 1998; Meck, 2006). However, whether dynamics in striatal activity can explain the perceptual performance of behaving subjects is unknown. To determine whether striatal ensembles drive subjects’ judgments of duration, we manipulated and recorded from striatal neurons in rats performing a duration categorization psychophysical task. Results To measure the duration sensitivity of subjects’ timing judgments, we trained rats to judge whether time intervals belonged to a long or short category (Gouvêa et al., 2014) (see Materials and methods; Figure 1a). At each self-initiated trial, two brief tones (interval onset, offset) were presented separated in time by an interval randomly selected from the set I = {0.6, 1.05, 1.26, 1.38, 1.62, 1.74, 1.95, 2.4} seconds. Judgments about interval duration were reported at two laterally located nose ports: choosing the left side was reinforced with water after intervals longer than 1.5 seconds (long stimuli), and the right side otherwise (short stimuli, Figure 1b). Animals were required to withhold choice until interval offset. Animals made virtually no errors when categorizing the easiest (i.e. shortest and longest) intervals, but categorization performance declined as intervals approached the 1.5 second categorical boundary (Figure 1c). Figure 1 with 1 supplement see all Download asset Open asset Rats judged interval durations as either long or short. (a) Rats triggered interval stimuli (i.e. two brief auditory tones separated by a silent interval of random duration) by inserting their snout into a central port. Following interval offset, animals reported their long vs short judgment at two lateral choice ports. Correct trials yielded a water reward, while incorrect or premature responses produced a white noise sound and a time out. Top view, high-speed video was acquired throughout task performance. (b) Reward contingency. (c) Averaged psychometric curves following bilateral muscimol or saline injections in dorsal striatum (mean ± standard deviation across session means, and logistic fit; n = 3 rats, 4 sessions each). Inset: slope of psychometric curves on consecutive saline and muscimol sessions. All raw data for Figure 1 can be found in Figure 1—source data 1. https://doi.org/10.7554/eLife.11386.003 Figure 1—source data 1 The. txt file contains trial by trials stimulus (Interval), choice (choiceLong), animal (Name), treatment (MuscimolDose) and session (Date). https://doi.org/10.7554/eLife.11386.004 Download elife-11386-fig1-data1-v2.txt We recorded action potentials during task performance (see Materials and methods), from populations of single striatal neurons targeting dorsal-central striatum, an area where manipulations produced timing deficits (Meck, 2006) (Figure 2a. For a reconstruction of striatal recording sites see Figure 2—figure supplement 1.). We observed that striatal neurons displayed diverse firing patterns, with different units firing at different times within the interval period (Figure 2b–d). Can such firing patterns support duration judgments? To determine whether and the degree to which individual neurons could contribute to duration judgments, for each trial, we counted spikes in the last 500 ms of the interval period and compared spike count distributions of short vs long stimulus trials using a receiver operating characteristic (ROC) analysis (see Materials and methods). We found that the majority of neurons (~57%) preferred either short or long stimuli (Figure 2e; short-preferring: n = 159/433, 36.7%; long-preferring: n = 87/433, 20.1%; permutation test, p<0.05). As expected, short-preferring neurons displayed higher firing on average prior to the 1.5 s category boundary, after which long-preferring neurons displayed higher firing (Figure 2f). These averaged activity patterns resemble the likelihood of receiving reward on a moment-by-moment by basis should the animal choose short or long (compare with reward contingency in Figure 1b). Such signals, previously observed in the parietal cortex of monkeys performing a similar timing task (Leon and Shadlen, 2003) and in the striatum in a value based decision task (Lau and Glimcher, 2008), are potentially useful for guiding choice. However, were animals’ judgments indeed guided by such signals, it should be possible to predict choices reported later in the trial using neural activity collected during interval stimuli. Indeed, in trials wherein a near boundary interval was judged as long, firing of the short (long) preferring subpopulation dropped (rose) faster, so that the two curves crossed before the 1.5 s boundary (Figure 2g). Conversely, in trials wherein the same interval was judged as short, the two curves evolved more slowly so that at the time of interval offset the short preferring subpopulation was still firing at a higher level and a crossing point had not yet been reached (Figure 2h). Figure 2 with 2 supplements see all Download asset Open asset Dynamics of striatal subpopulations predict duration judgments. (a) Psychometric function for neural recording sessions (mean ± standard deviation across sessions and logistic fit, n = 37 sessions from 3 rats). (b,c) Raster plot and peri-stimulus time histogram (PSTH) of two example cells for trials in which the longest stimulus interval (2.4 s) was presented. Time = 0 corresponds to stimulus onset. (d) Normalized PSTHs of all neurons for trials in which the longest stimulus interval was presented. Arrowheads indicate cells shown in (b,c). Blue and red ticks indicate cells with significant short and long preferences, respectively. (e) Histogram of preference indices. Blue and red outlines indicate subpopulations with significant short and long preferences, respectively. (f) Averaged, normalized PSTH of the two subpopulations outlined in (e) for trials in which the longest stimulus interval was presented (mean ± SEM). (g) Same as in (f), for trials in which a near-boundary stimulus interval (1.62 s) was judged as long. For comparison, curves shown in (f) are reproduced as a watermark. (h) same as (g) for trials in which the stimulus was judged as short. For single subjects, see Figure 2—figure supplement 2 . Behavior and neural spike count data for Figure 2 and Figure 2—figure supplements 1 and 2 can be found in Figure 2—source data 1. https://doi.org/10.7554/eLife.11386.006 Figure 2—source data 1 Folder with raw data for Figures 2–4. Each file in the source data folder is named “‘_"’, and contains an 8x2 cell "‘spikeCounts"’. For each of the eight stimuli and two choices, "‘spikeCounts"’ contains a matrix of spike counts with dimensions:number of neurons x number of trials x number of time steps. The stimuli (delay to the second tone in seconds) and choices are ordered as follows:STIM_SET = [0.60 1.05 1.26 1.38 1.62 1.74 1.95 2.4];CHOICES = ['incorrect' 'correct'];The spike counts are given from 500 ms prior to trial initiation up to 500 ms after the time of the second tone and were binned in 2 ms wide bins.This folder also includes underlying data for Figures 3 and 4. https://doi.org/10.7554/eLife.11386.007 Download elife-11386-fig2-data1-v2.zip The observation of large proportions of short- and long-preferring neurons whose dynamics predicted choice is evidence that duration judgments are guided by the state of striatal populations. Might the information afforded by ensembles of striatal neurons account for the pattern of subjects’ judgments across all stimuli? To test this hypothesis, we compared session to session fluctuations in behavioral performance with the separability of activity states of simultaneously recorded ensembles at the offset of short as compared to long intervals. Briefly, for each trial in a session we characterized neural population state as a vector r = (r1, r2,... , rN), where rn is the number of spikes fired by neuron n ∈ [1,N] within the last 500 ms of the interval period. Next, for each session we found the linear discriminant that best separated population state vectors according to whether they came from a long or a short interval trial (Figure 3a; see Materials and methods). A threshold placed along the linear discriminant was then used as a decision rule (black line in Figure 3a) to generate a ‘neural duration judgment’ for each trial. This procedure allowed us to obtain, for each session, a quantitative description of how well simultaneously recorded neurons could categorize stimuli, i.e., a neurometric function comparable to the behavioral psychometric function (Figure 3b). Consistent with duration information being encoded at the population level, we found that for sessions in which greater numbers of neurons were recorded simultaneously (i.e. upper tercile of sessions with regard to population size) psychometric and neurometric performances were similar and strongly correlated (r2 = 0.76, p<0.001; Figure 3c). These results demonstrate that a read out of stimulus category from even modestly-sized ensembles of striatal neurons was in many cases sufficient to explain the pattern of duration judgments produced by behaving subjects. Figure 3 with 3 supplements see all Download asset Open asset Simultaneously recorded population state at interval offset can explain behavioral performance. (a) Low dimensional representation of population state at interval offset for one example session. Black line is the decision rule (see text). (b) Example psychometric, neurometric and videometric curves for the same session as in (a). (c) Slopes of psychometric and neurometric curves for all sessions. Color indicates terciles of population size. Highlighted data point corresponds to the session in (a-b). Inset: regression slope of neurometric and videometric curves for sessions in the upper tercile. See Figure 3—figure supplement 2 for psychometric-videometric comparison at interval offset and choice. (d) Performance of an ideal observer analysis in predicting stimulus category, applied to neural (orange) and video (blue) data obtained at different times relative to interval offset. Thin lines corresponds to individual sessions. Thick lines are averages. (e) The orange and blue curves (thin lines in panel (d)) for corresponding sessions were regressed against each other at different time shifts. The regression R2 values for each session are shown in thin grey lines. The average over all sessions is shown in black. Sizes of black squares indicate the number of sessions with significant positive correlations (largest squares at 200 and 100 ms correspond to 5 sessions and smallest one at -200 ms to 1, out of a total of 8 sessions). (f) Psychometric curves constructed from trials separated according to whether the population state at stimulus offset had advanced more or less along the mean trajectory. Color indicates terciles shown in (g). (g) Distributions of projection on normalized mean trajectory for all trials for each stimulus are shown (stimuli color coded as in [a]). The equal area bins shown correspond to the groups of trials used for constructing the three psychometric curves shown in panel (f). Data in f-g are from rat Bertrand. See Figure 3—figure supplement 3 for the remaining two subjects. https://doi.org/10.7554/eLife.11386.010 It has been previously reported that duration judgments could be predicted by animals’ ongoing behavior during the interval period (Gouvêa et al., 2014; Matthews and Lerer, 1987; Killeen and Fetterman, 1988; Fetterman et al., 1998; Machado, 1997; Machado and Keen, 2003). In addition, it is well known that striatal neurons can fire around movements (Alexander and Crutcher, 1990; Jin and Costa, 2010). Could the categorization performance of striatal ensembles reflect activity related to movements the animal might be making during the task? To test to what degree ongoing behavior could explain the categorization performance of striatal neural activity, we applied an analogous classification analysis to video images taken of the animal just before interval offset (see Materials and methods). We found that our ability to categorize intervals using video frames was consistently poorer as compared to neural data collected at the analogous time periods during the task (Figure 3b, inset in Figure 3c, Video 1, Figure 3—figure supplement 1, Figure 3—figure supplement 2a). In contrast, we were able to categorize stimuli as well as the animal using video frames taken at the point when animals expressed their choice at one of the reward ports (Figure 3—figure supplement 2b). Furthermore, movement related responses in the striatum are known to occur both pre- and post-movement onset, much later than in other motor areas such as pre-motor and motor cortex (Alexander and Crutcher, 1990). Thus, if purely movement-related activity were responsible for the categorization performance of striatal ensembles, we would expect ensemble performance to display a similar time course to that of video frames. Applying the same analyses at multiple points in time ranging from 500 ms preceding to 500 ms following stimulus offset revealed a strikingly different profile of categorization performance for video frames as compared to neural ensembles (Figure 3d–e). Specifically, the time course of duration categorization by neural ensembles was best correlated with the duration categorization by video frames when using spikes collected between 400 ms and 200 ms preceding a reference video frame. These indicate that the categorization performance of striatal neurons was not simply related to the immediate sensorimotor state of the animal, and instead likely reflects that striatal neurons encode an internal neural representation of the state of animals’ categorical decisions. Video 1 Download asset This video cannot be played in place because your browser does support HTML5 video. You may still download the video for offline viewing. Download as MPEG-4 Download as WebM Download as Ogg Video clips from the entire stimulus period do not show a consistent separation between short and long stimulus trials. Superimposed video clips (thresholded and background subtracted) for each stimulus type and for the corresponding stimulus duration (same color conventions as in Figure 3a and Figure 3—figure supplement 1). Red stop marks signal the end of each video and the corresponding stimulus duration. https://doi.org/10.7554/eLife.11386.014 We have shown thus far that categorical information about interval duration contained in the firing of striatal populations at the time of stimulus offset can explain the precision of animals’ judgments about duration. However, in the task employed here, categorical judgments must be derived from a continuously evolving decision variable that represents how much time has elapsed since the onset of the stimulus. As indicated by the diversity of firing patterns (Figure 2d), the state of population activity evolved continuously during interval stimuli (Figure 3g, Figure 4a, Figure 3—figure supplement 3b), a feature not captured by binary classification. Might trial to trial variations in population state predict the apparent speed of animals’ internal representation of elapsed time? To test this possibility, we performed two additional analyses. Figure 4 with 2 supplements see all Download asset Open asset Smoothly changing population state encodes elapsing time in accordance with perceptual report for a long stimulus. (a) Low dimensional representation of population state during entire interval period of correct trials. Line colors indicate interval duration (warmer colors are longer intervals, as in Figure 3a). Dots are placed at the interval offset end, and their color indicates judgment (blue: short; red: long). (b–g) Population state and decoded time for a single long, near boundary stimulus interval (1.62 s). (b) Yellow curve is same as in (a). Red dots are 6 time points evenly spaced between interval onset and offset. Blue dots are projections of population state during short judgment trials. Grey lines link population states at equivalent time points. (c) Average cumulative distance travelled in full neural space along trajectory represented in (b) on long versus short judgment trials. (d) Posterior probability of time given population state at the time points indicated in (b), averaged within trials of each judgment type. (e,f) Same as (d) for the entire interval period. (g) Difference between posteriors for long and short judgment trials. Arrowheads indicate same time points used in (b,d). n = 433 neurons from 3 rats. See Figure 4—figure supplement 1 for a different near boundary stimulus, and Figure 4—figure supplement 2 for data from individual subjects. https://doi.org/10.7554/eLife.11386.015 First, we projected the state of simultaneously recorded neuronal populations at stimulus offset in individual trials onto the mean trajectory within each session, noted the fraction of the mean trajectory traversed for each trial, and pooled the data for each stimulus over all sessions within a given subject. Indeed, when population state at stimulus offset had advanced relatively more or less along the mean trajectory, animals were more likely to judge intervals as long or short respectively (Figure 3f–g, Figure 3—figure supplement 3). This effect was observed most consistently for interval stimuli that were closer to the category boundary, and thus variations in projected population state led to horizontal shifts in the psychometric curves (see Materials and methods). These data are consistent with striatal population state encoding a representation of elapsed time that is used by animals to determine categorical judgments. Indeed, such a pattern of population activity has been proposed as a suitable neural code for elapsing time (Buonomano, 2014; Buonomano and Merzenich, 1995). However, if such a representation encodes elapsed time, and not only subjects’ judgments in this task, neural activity should continuously traverse a non-repeating trajectory in state space in a manner that predicts duration judgments during presentation of particular stimuli. Indeed, even in a low dimensional projection of population activity, we found that network state ran ahead or behind depending on whether the animal judged a near boundary stimulus as long or short (Figure 4b–c, Figure 4—figure supplement 1b,c, Figure 4—figure supplement 2b,f,j). The correspondence between population trajectory and duration judgments further suggests that striatal dynamics may form an internal representation of elapsed time that informed categorical decisions about duration. To directly test this hypothesis, we focused on stimuli near the category boundary and decoded time from the population using a naive Bayes decoder and asked whether such a representation correlated with animals’ judgments, exhibiting choice probability (Britten et al., 1996). We found that decoded estimates of time ran faster or slower when animals judged a given stimulus as long or short, respectively (Figure 4d–g, Figure 4—figure supplement 1d-g, cross validated naive Bayes decoder; see Materials and methods). This indicates that striatal activity provides information about elapsing time, the continuously varying decision variable necessary to inform judgments in the task. Furthermore, if this information were read out and used to guide judgments, those judgments would match those of the rats. If the striatal activity we describe above directly supported task performance, manipulating the striatum should modify duration judgments. To test whether this was the case, we bilaterally injected the GABAa receptor agonist muscimol (see Materials and methods). As a result, the duration sensitivity of animals’ judgments dropped significantly as compared to interleaved saline control sessions (Figure 1c; psychometric slope on saline sessions = [1.03 1.20] vs on muscimol sessions = [0.28 0.67]; 95% confidence intervals), yet animals otherwise performed normally. These results, by demonstrating that duration categorization in this task was dependent on a normally functioning striatum, suggest that the neural signals we observed directly supported duration judgments. However, the possibility that muscimol infusions changed other functions important for task performance such as reward processing or memory for the mapping between time and choice can not be ruled out. Discussion Attempts to understand the neural mechanisms of time estimation have begun to focus on continuously evolving population dynamics as a general mechanism for time encoding across the brain (MacDonald et al., 2011; Mauk and Buonomano, 2004; Buonomano, 2014; Buonomano and Merzenich, 1995; Gershman et al., 2013). According to this view, time may be encoded by any reproducible pattern of activity across a population of neurons for as long as the pattern is continuously changing and non-repeating. However, no study to date has directly compared the speed of such “population clocks” with the duration judgments of the behaving subjects in which they are found. We show that as rats judged the duration of interval stimuli, striatal neurons displayed dynamics in firing rate that contained information about elapsed time. Furthermore, this information was sufficient to account for the animals’ perceptual decisions, and was not accompanied by systematic differences in outwardly expressed behavior over time. Combined with the observation that striatal inactivation caused a specific decrement in timing performance, these data suggest that striatal dynamics form a central neural representation of time that guides animals’ decisions about duration. Such a coding mechanism in the striatum is well situated to inform the appropriate selection of actions through downstream circuitry involving the globus pallidus, substantia nigra, and various extrinsic connections between the basal ganglia and brainstem, thalamic, and cortical motor areas (Steiner and Tseng, 2010). An intriguing question for future studies is how the striatal dynamics we observed during the interval discrimination task are generated. Neurons in multiple cortical layers spread across the entire cortical mantle, as well as thalamic, pallidal, and neuromodulatory populations provide input to the striatum. While time coding has been assessed in some of these populations, a careful analysis of simultaneously recorded populations that might uncover causal relationships between signals in multiple brain areas has not been carried out. Furthermore, local striatal circuitry may also play a role in shaping dynamics. However, the coding properties tested here could be tested in other brain areas where timing signals have been identified such as the hippocampus (Pastalkova et al., 2008; MacDonald et al., 2011), medial prefrontal (Kim et al., 2013; Xu et al., 2014), parietal (Leon and Shadlen, 2003; Janssen and Shadlen, 2005) and motor (Lebedev et al., 2008) cortices, and the cerebellum (Mauk and Buonomano, 2004), among others. By comparing the signals recorded simultaneously in multiple brain areas during time estimation tasks, it should be possible to identify signatures of functional interaction between brain areas where they exist. Such an approach promises to elucidate where and how time information encoded at the population level is used by the brain to support the myriad time-dependent functions we and other organisms rely on for survival. Materials and methods Subject Request a detailed protocol Six male Long-Evans hooded rats (Rattus norvegicus) between the ages of 6 and 24 months were used for this study. Three rats were used for neural recordings and three rats for pharmacological manipulations. All experiments were in accordance with the European Union Directive 86/609/EEC and approved by the Portuguese Veterinary General Board (Direcção-Geral de Veterinária, project approval 014303 - 0420/000/000/2011). Behavior Request a detailed protocol Rats were trained to perform a previously described two-alternative forced-choice timing task (Gouvêa et al., 2014). Briefly, animals had to categorize time intervals as either long or short by making left/right choices. For each session the animals were placed in a custom made behavioral box containing 3 nose ports and a speaker. Each trial was self-initiated by entry into the central nose port and was followed by a pair of brief auditory tones (square pulses at 7,500 Hz, 150 ms) separated by an interval selected randomly out of 8 possible durations (0.6, 1.05, 1.26, 1.38, 1.62, 1.74, 1.95 and 2.4 s). Judgments were reported at two laterally located nose ports. Left responses were reinforced with a drop of water (solenoid valves, Lee Company) after intervals longer than 1.5 seconds, and right responses otherwise. Incorrect responses were punished with a brief white noise burst (150 ms) and a time out (10 s). High speed video (120 fps) was collected from above during task performance. Psychometric functions were fitted using the following two

To guide behavior and learn from its consequences, the brain must represent time over many scales. Yet, the neural signals used to encode time in the seconds-to-minute range are not known. The striatum is a major input area of the basal ganglia associated with learning and motor function. Previous studies have also shown that the striatum is necessary for normal timing behavior. To address how striatal signals might be involved in timing, we recorded from striatal neurons in rats performing an interval timing task. We found that neurons fired at delays spanning tens of seconds and that this pattern of responding reflected the interaction between time and the animals' ongoing sensorimotor state. Surprisingly, cells rescaled responses in time when intervals changed, indicating that striatal populations encoded relative time. Moreover, time estimates decoded from activity predicted timing behavior as animals adjusted to new intervals, and disrupting striatal function led to a decrease in timing performance. These results suggest that striatal activity forms a scalable population code for time, providing timing signals that animals use to guide their actions.

The design of modern scientific experiments requires the control and monitoring of many different data streams. However, the serial execution of programming instructions in a computer makes it a challenge to develop software that can deal with the asynchronous, parallel nature of scientific data. Here we present Bonsai, a modular, high-performance, open-source visual programming framework for the acquisition and online processing of data streams. We describe Bonsai's core principles and architecture and demonstrate how it allows for the rapid and flexible prototyping of integrated experimental designs in neuroscience. We specifically highlight some applications that require the combination of many different hardware and software components, including video tracking of behavior, electrophysiology and closed-loop control of stimulation.

The striatum is an input structure of the basal ganglia implicated in several time- dependent functions including reinforcement learning, decision making, and interval timing. To determine whether striatal ensembles drive subjects' judgments of duration, we manipulated and recorded from striatal neurons in rats performing a duration categorization psychophysical task. We found that the dynamics of striatal neurons predicted duration judgments, and that simultaneously recorded ensembles could judge duration as well as the animal. Furthermore, striatal neurons were necessary for duration judgments, as muscimol infusions produced a specific impairment in animals' duration sensitivity. Lastly, we show that time as encoded by striatal populations ran faster or slower when rats judged a duration as longer or shorter, respectively. These results demonstrate that the speed with which striatal population state changes supports the fundamental ability of animals to judge the passage of time.

The striatum is an input structure of the basal ganglia implicated in several time-dependent functions including reinforcement learning, decision making, and interval timing. To determine whether striatal ensembles drive subjects' judgments of duration, we manipulated and recorded from striatal neurons in rats performing a duration categorization psychophysical task. We found that the dynamics of striatal neurons predicted duration judgments, and that simultaneously recorded ensembles could judge duration as well as the animal. Furthermore, striatal neurons were necessary for duration judgments, as muscimol infusions produced a specific impairment in animals' duration sensitivity. Lastly, we show that time as encoded by striatal populations ran faster or slower when rats judged a duration as longer or shorter, respectively. These results demonstrate that the speed with which striatal population state changes supports the fundamental ability of animals to judge the passage of time.

The ability to estimate the passage of time is essential for adaptive behavior in complex environments. Yet, it is not known how the brain encodes time over the durations necessary to explain animal behavior. Under temporally structured reinforcement schedules, animals tend to develop temporally structured behavior, and interval timing has been suggested to be accomplished by learning sequences of behavioral states. If this is true, trial to trial fluctuations in behavioral sequences should be predictive of fluctuations in time estimation. We trained rodents in an duration categorization task while continuously monitoring their behavior with a high speed camera. Animals developed highly reproducible behavioral sequences during the interval being timed. Moreover, those sequences were often predictive of perceptual report from early in the trial, providing support to the idea that animals may use learned behavioral patterns to estimate the duration of time intervals. To better resolve the issue, we propose that continuous and simultaneous behavioral and neural monitoring will enable identification of neural activity related to time perception that is not explained by ongoing behavior.