Publications

Journal Publications

2021

Jian-You Lin, Narendra Mukherjee, Max J Bernstein, and Donald B Katz. Perturbation of amygdala-cortical projections reduces ensemble coherence of palatability coding in gustatory cortex. eLife, 10:e65766, 2021.
[abstract▼] [full text] [BibTeX▼]

Taste palatability is centrally involved in consumption decisions - we ingest foods that taste good and reject those that don't. Gustatory cortex (GC) and basolateral amygdala (BLA) almost certainly work together to mediate palatability-driven behavior, but the precise nature of their interplay during taste decision-making is still unknown. Here, we take a step toward filling this gap in our knowledge, by investigating the specific role that activity in the BLA->GC pathway plays in the emergence of palatability-related firing in GC response dynamics (which influence consumption decisions). We implanted electrode/optical-fiber probes in virally-prepared female Long-Evans rats, such that we could optogenetically hyperpolarize BLA->GC axons, perturbing activity in these axons without affecting BLA and GC somas, while recording GC neural responses to intra-oral presentations of a diverse taste battery. This inter-regional axonal perturbation strongly altered GC taste responses, but despite the laser illumination being tonic for the first 2s that the taste was on the tongue, the alterations were far from monolithic: rather than changing all moments of the response equally, or causing a simple exponential decay of changes, the perturbation was most strongly felt at the onset times of previously-described response epochs; furthermore, the effect was epoch-specific - perturbations had little impact on the amount of taste identity information in the "middle epoch" of the responses, but reduced evidence of palatability-related activity in the "late-epoch." Finally, BLA->GC axon inhibition affected the nature of the epochal dynamics themselves, such that the normal abruptness of the behaviorally-relevant ensemble transitions into the palatability-related epoch was greatly diminished. These results suggest that BLA "organizes" behavior-related GC taste dynamics.
```
@article{Lin2020amygdala,
    author = "Lin, Jian-You and Mukherjee, Narendra and Bernstein, Max J and Katz, Donald B",
    title = "Perturbation of amygdala-cortical projections reduces ensemble coherence of palatability coding in gustatory cortex",
    journal = "eLife",
    volume = "10",
    pages = "e65766",
    year = "2021",
    publisher = "eLife Sciences Publications Limited"
}
```

2019

Narendra Mukherjee, Joseph Wachutka, and Donald B Katz. Impact of precisely-timed inhibition of gustatory cortex on taste behavior depends on single-trial ensemble dynamics. eLife, 8:e45968, 2019.
[abstract▼] [full text] [BibTeX▼]

Sensation and action are necessarily coupled during stimulus perception – while tasting, for instance, perception happens while an animal decides to expel or swallow the substance in the mouth (the former via a behavior known as ‘gaping’). Taste responses in the rodent gustatory cortex (GC) span this sensorimotor divide, progressing through firing-rate epochs that culminate in the emergence of action-related firing. Population analyses reveal this emergence to be a sudden, coherent and variably-timed ensemble transition that reliably precedes gaping onset by 0.2–0.3s. Here, we tested whether this transition drives gaping, by delivering 0.5s GC perturbations in tasting trials. Perturbations significantly delayed gaping, but only when they preceded the action-related transition - thus, the same perturbation impacted behavior or not, depending on the transition latency in that particular trial. Our results suggest a distributed attractor network model of taste processing, and a dynamical role for cortex in driving motor behavior.
```
@article{mukherjee2019impact,
    author = "Mukherjee, Narendra and Wachutka, Joseph and Katz, Donald B",
    title = "Impact of precisely-timed inhibition of gustatory cortex on taste behavior depends on single-trial ensemble dynamics",
    journal = "eLife",
    volume = "8",
    pages = "e45968",
    year = "2019",
    publisher = "eLife Sciences Publications Limited"
}
```
Narendra Mukherjee. Behaviorally Relevant Sensory Cortical Population Dynamics in the Rodent Taste System. PhD thesis, Brandeis University, 2019.
[abstract▼] [full text] [BibTeX▼]

In the studies that are part of this dissertation, we investigated the role of primary sensory taste cortex in generation of behavioral responses to taste stimuli. Starting with a general description of prior work and formulation of a roadmap for the thesis, each chapter describes a separate investigation of the connection between cortical processing of sensory stimuli, appropriately-timed behavioral outputs and associative learning based on sensory experiences. The chapters each start with their own abstract, specific to the study being reported.
```
@phdthesis{mukherjee2019behaviorally,
    author = "Mukherjee, Narendra",
    title = "Behaviorally Relevant Sensory Cortical Population Dynamics in the Rodent Taste System",
    year = "2019",
    school = "Brandeis University"
}
```
David Levitan, Jian-You Lin, Joseph Wachutka, Narendra Mukherjee, Sacha B Nelson, and Donald B Katz. Single and population coding of taste in the gustatory-cortex of awake mice. Journal of Neurophysiology, 2019.
[abstract▼] [full text] [BibTeX▼]

Electrophysiological analysis has revealed much about the broad coding and neural ensemble dynamics that characterize gustatory cortical (GC) taste processing in awake rats and about how these dynamics relate to behavior. With regard to mice, however, data concerning cortical taste coding have largely been restricted to imaging, a technique that reveals average levels of neural responsiveness but that (currently) lacks the temporal sensitivity necessary for evaluation of fast response dynamics; furthermore, the few extant studies have thus far failed to provide consensus on basic features of coding. We have recorded the spiking activity of ensembles of GC neurons while presenting representatives of the basic taste modalities (sweet, salty, sour, and bitter) to awake mice. Our first central result is the identification of similarities between rat and mouse taste processing: most mouse GC neurons (~66%) responded distinctly to multiple (3–4) tastes; temporal coding analyses further reveal, for the first time, that single mouse GC neurons sequentially code taste identity and palatability, the latter responses emerging ~0.5 s after the former, with whole GC ensembles transitioning suddenly and coherently from coding taste identity to coding taste palatability. The second finding is that spatial location plays very little role in any aspect of taste responses: neither between- (anterior-posterior) nor within-mouse (dorsal-ventral) mapping revealed anatomic regions with narrow or temporally simple taste responses. These data confirm recent results showing that mouse cortical taste responses are not “gustotopic” but also go beyond these imaging results to show that mice process tastes through time.
```
@article{levitan2019single,
    author = "Levitan, David and Lin, Jian-You and Wachutka, Joseph and Mukherjee, Narendra and Nelson, Sacha B and Katz, Donald B",
    title = "Single and population coding of taste in the gustatory-cortex of awake mice",
    journal = "Journal of Neurophysiology",
    year = "2019"
}
```

2018

Veronica L Flores, Tamar Parmet, Narendra Mukherjee, Sacha Nelson, Donald B Katz, and David Levitan. The role of the gustatory cortex in incidental experience-evoked enhancement of later taste learning. Learning & Memory, 25(11):587–600, 2018. doi:10.1101/lm.048181.118.
[abstract▼] [full text] [BibTeX▼]

The strength of learned associations between pairs of stimuli is affected by multiple factors, the most extensively studied of which is prior experience with the stimuli themselves. In contrast, little data is available regarding how experience with “incidental” stimuli (independent of any conditioning situation) impacts later learning. This lack of research is striking given the importance of incidental experience to survival. We have recently begun to fill this void using conditioned taste aversion (CTA), wherein an animal learns to avoid a taste that has been associated with malaise. We previously demonstrated that incidental exposure to salty and sour tastes (taste preexposure—TPE) enhances aversions learned later to sucrose. Here, we investigate the neurobiology underlying this phenomenon. First, we use immediate early gene (c-Fos) expression to identify gustatory cortex (GC) as a site at which TPE specifically increases the neural activation caused by taste-malaise pairing (i.e., TPE did not change c-Fos induced by either stimulus in isolation). Next, we use site-specific infection with the optical silencer Archaerhodopsin-T to show that GC inactivation during TPE inhibits the expected enhancements of both learning and CTA-related c-Fos expression, a full day later. Thus, we conclude that GC is almost certainly a vital part of the circuit that integrates incidental experience into later associative learning.
```
@article{flores2018role,
    author = "Flores, Veronica L and Parmet, Tamar and Mukherjee, Narendra and Nelson, Sacha and Katz, Donald B and Levitan, David",
    title = "The role of the gustatory cortex in incidental experience-evoked enhancement of later taste learning",
    journal = "Learning \\& Memory",
    volume = "25",
    number = "11",
    pages = "587--600",
    year = "2018",
    publisher = "Cold Spring Harbor Lab",
    doi = "10.1101/lm.048181.118"
}
```

2016

Brian F Sadacca, Narendra Mukherjee, Tony Vladusich, Jennifer X Li, Donald B Katz, and Paul Miller. The behavioral relevance of cortical neural ensemble responses emerges suddenly. Journal of Neuroscience, 36(3):655–669, 2016. doi:10.1523/JNEUROSCI.2265-15.2016.
[abstract▼] [full text] [BibTeX▼]

Whereas many laboratory-studied decisions involve a highly trained animal identifying an ambiguous stimulus, many naturalistic decisions do not. Consumption decisions, for instance, involve determining whether to eject or consume an already identified stimulus in the mouth and are decisions that can be made without training. By standard analyses, rodent cortical single-neuron taste responses come to predict such consumption decisions across the 500 ms preceding the consumption or rejection itself; decision-related firing emerges well after stimulus identification. Analyzing single-trial ensemble activity using hidden Markov models, we show these decision-related cortical responses to be part of a reliable sequence of states (each defined by the firing rates within the ensemble) separated by brief state-to-state transitions, the latencies of which vary widely between trials. When we aligned data to the onset of the (late-appearing) state that dominates during the time period in which single-neuron firing is correlated to taste palatability, the apparent ramp in stimulus-aligned choice-related firing was shown to be a much more precipitous coherent jump. This jump in choice-related firing resembled a step function more than it did the output of a standard (ramping) decision-making model, and provided a robust prediction of decision latency in single trials. Together, these results demonstrate that activity related to naturalistic consumption decisions emerges nearly instantaneously in cortical ensembles.
```
@article{sadacca2016behavioral,
    author = "Sadacca, Brian F and Mukherjee, Narendra and Vladusich, Tony and Li, Jennifer X and Katz, Donald B and Miller, Paul",
    title = "The behavioral relevance of cortical neural ensemble responses emerges suddenly",
    journal = "Journal of Neuroscience",
    volume = "36",
    number = "3",
    pages = "655--669",
    year = "2016",
    publisher = "Soc Neuroscience",
    doi = "10.1523/JNEUROSCI.2265-15.2016"
}
```

2013

Vishwanath Varma, Narendra Mukherjee, Nisha N Kannan, and Vijay Kumar Sharma. Strong (type 0) phase resetting of activity-rest rhythm in fruit flies, drosophila melanogaster, at low temperature. Journal of Biological Rhythms, 28(6):380–389, 2013. doi:10.1177/0748730413508922.
[abstract▼] [full text] [BibTeX▼]

Amplitude modulation in limit cycle models of circadian clocks has been previously formulated to explain the phenomenon of temperature compensation. These models propose that invariance of clock period (τ) with changing temperature is a result of the system traversing small or large limit cycles such that despite a decrease or an increase in the linear velocity of the clock owing to slowing down or speeding up of the underlying biochemical reactions, respectively, the angular velocity and, thus, the clock period remain constant. In addition, these models predict that phase resetting behavior of circadian clocks described by limit cycles of different amplitudes at low or high temperatures will be drastically different. More specifically, this class of models predicts that at low temperatures, circadian clocks will respond to perturbations by eliciting larger phase shifts by virtue of their smaller amplitude and vice versa. Here, we present the results of our tests of this prediction: We examined the nature of photic phase response curves (PRCs) and phase transition curves (PTCs) for the circadian clocks of 4 wild-type fruit fly Drosophila melanogaster populations at 3 different ambient temperatures (18, 25, and 29 °C). Interestingly, we observed that at the low temperature of 18 °C, fly clocks respond to light perturbations more strongly, eliciting strong (type 0) PRCs and PTCs, while at moderate (25 °C) and high (29 °C) temperatures the same stimuli evoke weak (type 1) responses. This pattern of strong and weak phase resetting at low and high temperatures, respectively, renders support for the limit cycle amplitude modulation model for temperature compensation of circadian clocks.
```
@article{varma2013strong,
    author = "Varma, Vishwanath and Mukherjee, Narendra and Kannan, Nisha N and Sharma, Vijay Kumar",
    title = "Strong (Type 0) phase resetting of activity-rest rhythm in fruit flies, Drosophila melanogaster, at low temperature",
    journal = "Journal of Biological Rhythms",
    volume = "28",
    number = "6",
    pages = "380--389",
    year = "2013",
    publisher = "Sage Publications Sage CA: Los Angeles, CA",
    doi = "10.1177/0748730413508922"
}
```

2012

Narendra Mukherjee, Nisha N Kannan, Pankaj Yadav, and Vijay Kumar Sharma. A model based on oscillatory threshold and build-up of a developmental substance explains gating of adult emergence in drosophila melanogaster. Journal of Experimental Biology, 215(17):2960–2968, 2012. doi:10.1242/jeb.071290.
[abstract▼] [full text] [BibTeX▼]

Adult emergence (eclosion) of fruit flies Drosophila melanogaster under constant laboratory conditions follows a circadian pattern with bouts of eclosion recurring at approximately 24 h intervals. Under periodic light:dark (LD) cycles, adults emerge only during a specific time of the day followed by little or no emergence for the rest of the day. This phenomenon is therefore equated to a gate of emergence that, when open, allows adults to emerge and when closed, no emergence takes place. In this study, we attempt to understand the mechanism underlying adult emergence rhythm in D. melanogaster using a model based on interplay between developmental and circadian clock systems. The model is composed of an oscillatory threshold of a substance that builds up during pre-adult development. Computer simulations based on this model enabled us to make specific predictions about the ‘gate width’ of the adult emergence rhythm under conditions of fast/slow pre-adult development and short/long circadian periods, which we subsequently tested empirically. The main predictions from the simulations are: (1) flies with faster development have greater gate width and vice versa, and (2) flies with faster circadian clocks have shorter gate width and vice versa. To empirically validate these predictions, we carried out experiments on D. melanogaster populations known to have fast/slow pre-adult development, short/long circadian periods and narrow/wide gate width. Additionally, we manipulated the rate of pre-adult development of the above flies by increasing/decreasing ambient temperature to further examine the influence of developmental rates on gate width of adult emergence rhythm by a complementary approach. The results show that gate width is greatly influenced by the duration of pre-adult development and the length of circadian cycles. This suggests that the adult emergence rhythm of D. melanogaster may be based on mechanisms involving oscillatory threshold and build-up of a developmental substance.
```
@article{mukherjee2012model,
    author = "Mukherjee, Narendra and Kannan, Nisha N and Yadav, Pankaj and Sharma, Vijay Kumar",
    title = "A model based on oscillatory threshold and build-up of a developmental substance explains gating of adult emergence in Drosophila melanogaster",
    journal = "Journal of Experimental Biology",
    volume = "215",
    number = "17",
    pages = "2960--2968",
    year = "2012",
    publisher = "The Company of Biologists Ltd",
    doi = "10.1242/jeb.071290"
}
```
Narendra Mukherjee. Optimality and courtship behaviour in zebrafish, danio rerio. Master's thesis, Indian Institute of Science Education and Research Kolkata, 2012.
[abstract▼] [full text] [BibTeX▼]

Behaviors involving repeated investments on the part of the actor and diminishing marginal returns from these investments are seen to be optimized by maximizing the rate of returns from the behavior. This idea, known as the marginal value theorem, was initially used to study foraging strategies in animals when food is available in patches in an environment. In this study, we adopt the marginal value theorem and utilize it to study courtship behavior in zebrafish Danio rerio (with males being equivalent to foraging animals and females being compared to patches of resources). Our application of the marginal value theorem to courtship in zebrafish predicted that a male spends less time courting a female if the female density in the environment is high and vice-versa. The results of our experiments support these predictions and show that courtship time in zebrafish is indeed greatly influenced by the female density in the environment. Additionally, we went on to construct a return function in the form of the probability that a female being courted mates with the male courting her as a function of the time invested by the male in courtship (equivalently, the number of times the male physically interacts with the female). This return function, in a sense, gives an idea of the sexual response of a female as the investment in courtship from the side of a male increases and can potentially prove the existence of diminishing marginal returns from repeated investments in courtship on the part of the male.
```
@mastersthesis{mukherjee2012optimality,
    author = "Mukherjee, Narendra",
    title = "Optimality and Courtship Behaviour in Zebrafish, Danio Rerio",
    year = "2012",
    school = "Indian Institute of Science Education and Research Kolkata"
}
```
Nisha N Kannan, Narendra Mukherjee, and Vijay Kumar Sharma. Robustness of circadian timing systems evolves in the fruit fly drosophila melanogaster as a correlated response to selection for adult emergence in a narrow window of time. Chronobiology international, 29(10):1312–1328, 2012. doi:10.3109/07420528.2012.728550.
[abstract▼] [full text] [BibTeX▼]

Robustness is a fundamental property of biological timing systems that is likely to ensure their efficient functioning under a wide range of environmental conditions. Here we report the findings of our study aimed at examining robustness of circadian clocks in fruit fly Drosophila melanogaster populations selected to emerge as adults within a narrow window of time. Previously, we have reported that such flies display enhanced synchrony, accuracy, and precision in their adult emergence and activity/rest rhythms. Since it is expected that accurate and precise circadian clocks may confer enhanced stability in circadian time-keeping, we decided to examine robustness in circadian rhythms of flies from the selected populations by subjecting them to a variety of environmental conditions comprising of a range of photoperiods, light intensities, ambient temperatures, and constant darkness. The results revealed that adult emergence and activity/rest rhythms of flies from the selected stocks were more robust than controls, as they displayed enhanced stability under a wide variety of environmental conditions. These results suggest that selection for adult emergence within a narrow window of time results in the evolution of robustness in circadian timing systems of the fruit fly D. melanogaster.
```
@article{kannan2012robustness,
    author = "Kannan, Nisha N and Mukherjee, Narendra and Sharma, Vijay Kumar",
    title = "Robustness of circadian timing systems evolves in the fruit fly Drosophila melanogaster as a correlated response to selection for adult emergence in a narrow window of time",
    journal = "Chronobiology international",
    volume = "29",
    number = "10",
    pages = "1312--1328",
    year = "2012",
    publisher = "Taylor \\& Francis",
    doi = "10.3109/07420528.2012.728550"
}
```

Conference Publications

2021

Shrabastee Banerjee, Narendra Mukherjee, and M Amin Rahimian. Deep learning for simulation-based bayesian inference of hidden parameters in online reputation systems. In Workshop on Machine Learning for Consumers and Markets (MLCM) at the 27th ACM SIGKDD Conference on Knowledge Discovery & Data Mining (KDD). 2021.
[abstract▼] [full text] [BibTeX▼]

Online reputation systems are an essential component of electronic commerce platforms. However, despite their prevalence, online ratings are subject to selection biases since the decision to leave a rating depends on the specific consumer and their circumstances. There are a number of hidden parameters governing such selection biases but it is difficult to infer them directly from observed ratings given the complexity of reputation systems. In this work, we first propose a generative model that accounts for various behavioral phenomena behind online rating generation (e.g., cost to leaving a rating or herding). We then build upon recent advances in likelihood-free/simulation-based Bayesian inference using deep learning to infer the hidden parameters of the generative model in a scalable manner. The inference engine only takes the time series of ratings as input, and therefore can be used to model correlations of inferred cost parameters with various product features. As a preliminary proof of concept, we apply our model to a dataset of 450,000 product reviews submitted on Amazon.com. We find that the cost to leaving a negative review is much greater than a positive review, and a baseline level of bandwagon effects (in the form of herding) are present for the majority of products. Gaining a better understanding of the dynamics of reputation systems, namely, the conditions under which ratings are submitted, is crucial for marketers, brand managers, and designers of digital platforms, who can leverage this information to stimulate further reviews and better manage user generated content.
```
@inproceedings{banerjeekdd2021,
    author = "Banerjee, Shrabastee and Mukherjee, Narendra and Rahimian, M Amin",
    title = "Deep learning for simulation-based Bayesian inference of hidden parameters in online reputation systems",
    booktitle = "Workshop on Machine Learning for Consumers and Markets (MLCM) at the 27th ACM SIGKDD Conference on Knowledge Discovery \\& Data Mining (KDD)",
    year = "2021"
}
```

2017

Narendra Mukherjee, Joseph Wachutka, and Donald B Katz. Python meets systems neuroscience: affordable, scalable and open-source electrophysiology in awake, behaving rodents. In Proceedings of the 16th Python in Science Conference, 98 – 105. 2017. doi:10.25080/shinma-7f4c6e7-00e.
[abstract▼] [full text] [BibTeX▼]

In-vivo electrophysiology, the recording of neurons in the brains of awake, behaving animals, is currently undergoing paradigm shifts. There is a push towards moving to open-source technologies that can: 1) be adjusted to specific experiments; 2) be shared with ease; and 3) more affordably record from larger numbers of electrodes simultaneously. Here we describe our construction of a system that satisfies these three desirable properties using the scientific Python stack and Linux. Using a Raspberry Pi to control experimental paradigms, we build a completely open-source, HDF5-based analysis (spike sorting) toolkit in Python. This toolkit can be easily parallelized and scales to incorporate increasing electrode counts and longer recordings. Our rig costs about \$5000, an order of magnitude less than many comparable commercially available electrophysiology systems.
```
@inproceedings{narendra_mukherjee-proc-scipy-2017,
    author = "Mukherjee, Narendra and Wachutka, Joseph and Katz, Donald B",
    title = "{P}ython meets systems neuroscience: affordable, scalable and open-source electrophysiology in awake, behaving rodents",
    booktitle = "{P}roceedings of the 16th {P}ython in {S}cience {C}onference",
    pages = "98 - 105",
    year = "2017",
    doi = "10.25080/shinma-7f4c6e7-00e"
}
```