A universal probabilistic spike count model reveals ongoing modulation of neural variability

Liu, David; Lengyel, Máté

doi:10.1101/2021.06.27.450063

Cited by 7 publications

(18 citation statements)

References 85 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Note that in our framework, estimation of the neural activities (firing rates) does not rely on selecting a temporal bin size. This is unlike most previous state-of-the-art works [Liu and Lengyel 2021] where the output firing rates are substantially affected by the choice of bin size. Next, examining the behavioural predictions of our model, the neural activities presented in Figure 4 show that sooner responses (with high probabilities) are strongly related to the peak intensities in VISp.…”

Section: Methodsmentioning

confidence: 75%

“…Here, the total spike counts for all trials are calculated and concatenated for count windows of 5ms as inputs to the GLM model. NHPoisson Model [Cebrian 2015] : NHPoisson is a method for the modelling non homogeneous Poisson processes in time estimating maximum likelihood. The model is based on formulating the intensity as a function of time-dependent covariates. Universal Count Model [Liu and Lengyel 2021] : This model builds on sparse Gaussian processes (GP) to capture arbitrary spike count distributions flexibly relying on both observed and latent covariates. It uses scalable variational inference and can jointly infer the covariate-to-spike count distribution mappings and latent trajectories.…”

Section: Resultsmentioning

confidence: 99%

“…The model is based on formulating the intensity as a function of time-dependent covariates. • Universal Count Model [Liu and Lengyel 2021]: This model builds on sparse Gaussian processes (GP) to capture arbitrary spike count distributions flexibly relying on both observed and latent covariates. It uses scalable variational inference and can jointly infer the covariate-to-spike count distribution mappings and latent trajectories.…”

Section: Baseline Methodsmentioning

confidence: 99%

“…This universal probabilistic spike count model uses sparse Gaussian processes to derive spike count distributions. It consists of C Gaussian process (GP) priors, a basis expansion, and a linear-softmax mapping [Liu and Lengyel 2021]. Using 1 ms time bins, the spike counts of all trials associated with each neuron (in the Steinmetz dataset) or the neuronal population (in the synthetic dataset) are calculated and concatenated together.…”

Section: Supplementary Materialsmentioning

confidence: 99%

“…In this model, a 10-fold cross validation is utilized. The remainder of the parameters and model settings are tuned according to the suggestions in [Liu and Lengyel 2021].…”

Section: Supplementary Materialsmentioning

confidence: 99%

See 4 more Smart Citations

Neural Network Poisson Models for Behavioural and Neural Spike Train Data

Dezfouli

Nock

Arabzadeh

et al. 2020

Preprint

View full text Add to dashboard Cite

It is now possible to monitor the activity of a large number of neurons across the brain as animals perform behavioural tasks. A primary aim for modeling is to reveal (i) how sensory inputs are represented in neural activities and (ii) how these representations translate into behavioural responses. Predominant methods apply rather disjoint techniques to (i) and (ii); by contrast, we suggest an end-to-end model which jointly fits both behaviour and neural activities and tracks their covariabilities across trials using inferred noise correlations. Our model exploits recent developments of flexible, but tractable, neural network point-process models to characterize dependencies between stimuli, actions and neural data. We apply the framework to a dataset collected using Neuropixel probes in a visual discrimination task and analyse noise correlations to gain novel insights into the relationships between neural activities and behaviour.

show abstract

Section: Methodsmentioning

confidence: 75%

Section: Resultsmentioning

confidence: 99%

Section: Baseline Methodsmentioning

confidence: 99%

Section: Supplementary Materialsmentioning

confidence: 99%

“…In this model, a 10-fold cross validation is utilized. The remainder of the parameters and model settings are tuned according to the suggestions in [Liu and Lengyel 2021].…”

Section: Supplementary Materialsmentioning

confidence: 99%

See 3 more Smart Citations

Neural Network Poisson Models for Behavioural and Neural Spike Train Data

Dezfouli

Nock

Arabzadeh

et al. 2020

Preprint

View full text Add to dashboard Cite

show abstract

Scalable Bayesian GPFA with automatic relevance determination and discrete noise models

Jensen

Kao

Stone

et al. 2021

Preprint

View full text Add to dashboard Cite

Latent variable models are ubiquitous in the exploratory analysis of neural population recordings, where they allow researchers to summarize the activity of large populations of neurons in lower dimensional 'latent' spaces. Existing methods can generally be categorized into (i) Bayesian methods that facilitate flexible incorporation of prior knowledge and uncertainty estimation, but which typically do not scale to large datasets; and (ii) highly parameterized methods without explicit priors that scale better but often struggle in the low-data regime. Here, we bridge this gap by developing a fully Bayesian yet scalable version of Gaussian process factor analysis (bGPFA) which models neural data as arising from a set of inferred latent processes with a prior that encourages smoothness over time. Additionally, bGPFA uses automatic relevance determination to infer the dimensionality of neural activity directly from the training data during optimization. To enable the analysis of continuous recordings without trial structure, we introduce a novel variational inference strategy that scales near-linearly in time and also allows for non-Gaussian noise models more appropriate for electrophysiological recordings. We apply bGPFA to continuous recordings spanning 30 minutes with over 14 million data points from primate motor and somatosensory cortices during a self-paced reaching task. We show that neural activity progresses from an initial state at target onset to a reach-specific preparatory state well before movement onset. The distance between these initial and preparatory latent states is predictive of reaction times across reaches, suggesting that such preparatory dynamics have behavioral relevance despite the lack of externally imposed delay periods. Additionally, bGPFA discovers latent processes that evolve over slow timescales on the order of several seconds and contain complementary information about reaction time. These timescales are longer than those revealed by methods which focus on individual movement epochs and may reflect fluctuations in e.g. task engagement.

show abstract

Beyond the Euclidean brain: inferring non-Euclidean latent trajectories from spike trains

Jensen

Liu

Kao

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

Neuroscience faces a growing need for scalable data analysis methods that reduce the dimensionality of population recordings yet retain key aspects of the computation or behaviour. To extract interpretable latent trajectories from neural data, it is critical to embrace the inherent topology of the features of interest: head direction evolves on a ring or torus, 3D body rotations on the special orthogonal group, and navigation is best described in the intrinsic coordinates of the environment. Accordingly, we recently proposed the manifold Gaussian process latent variable model (mGPLVM) to simultaneously infer latent representations on non-Euclidean manifolds and how neurons are tuned to these representations. This probabilistic method generalizes previous Euclidean models and allows principled selection between candidate latent topologies. While powerful, mGPLVM makes two unjustified approximations that limit its practical applicability to neural datasets. First, consecutive latent states are assumed independent a priori, whereas behaviour is continuous in time. Second, its Gaussian noise model is inappropriate for positive integer spike counts. Previous work in Euclidean LVMs such as GPFA has shown significant improvements in performance when modeling such features appropriately (Jensen et al., 2021). Here, we extend mGPLVM by incorporating temporally continuous priors over latent states and flexible count-based noise models. This improves inference on synthetic data, avoiding negative spike count predictions and discontinuous jumps in latent trajectories. On real data, we also mitigate these pathologies while improving model fit compared to the original mGPLVM formulation. In summary, our extended mGPLVM provides a widely applicable tool for inferring (non-)Euclidean neural representations from large-scale, heterogeneous population recordings. We provide an efficient implementation in python, relying on recent advances in approximate inference to e.g. fit 10,000 time bins of recording for 100 neurons in five minutes on a single GPU.

show abstract

A universal probabilistic spike count model reveals ongoing modulation of neural variability

Cited by 7 publications

References 85 publications

Neural Network Poisson Models for Behavioural and Neural Spike Train Data

Neural Network Poisson Models for Behavioural and Neural Spike Train Data

Scalable Bayesian GPFA with automatic relevance determination and discrete noise models

Beyond the Euclidean brain: inferring non-Euclidean latent trajectories from spike trains

Contact Info

Product

Resources

About