Seminars

The assumed data generating models (response distributions) of experimental or observational data in cognitive science have become increasingly complex over the past decades. This trend follows a revolution in model estimation methods and a drastic increase in computing power available to researchers. Today, higher-level cognitive functions can well be captured by and understood through computational cognitive models, a common example being drift diffusion models for decision processes. Such models are often expressed as the combination of two modeling layers. The first layer is the response distribution with corresponding distributional parameters tailored to the cognitive process under investigation. The second layer are latent models of the distributional parameters that capture how those parameters vary as a function of design, stimulus, or person characteristics, often in an additive manner. Such cognitive models can thus be understood as special cases of distributional regression models where multiple distributional parameters, rather than just a single centrality parameter, are predicted by additive models. Because of their complexity, distributional models are quite complicated to estimate, but recent advances in Bayesian estimation methods and corresponding software make them increasingly more feasible. In this talk, I will speak about the specification, estimation, and post-processing of Bayesian distributional regression models and how they can help to better understand cognitive processes.

Join the Department of Bioengineering on the 26th May at 9:00am for The 2021 Annual Bioengineering Lecture + Bioinspired Guidance, Navigation and Control Symposium. This year’s lecture speaker will be distinguished bioengineer and neuroscientist Professor Mandyam V. Srinivasan AM FRS, from the University of Queensland. Professor Srinivasan studies visual systems, particularly those of bees and birds. His research has revealed how flying insects negotiate narrow gaps, regulate the height and speed of flight, estimate distance flown, and orchestrate smooth landings. Apart from enhancing fundamental knowledge, these findings are leading to novel, biologically inspired approaches to the design of guidance systems for unmanned aerial vehicles with applications in the areas of surveillance, security and planetary exploration. Following Professor Srinivasan’s lecture will be the Bioinspired GNC Mini Symposium with guest speakers from Google Deepmind, Imperial College London, the University of Würzburg and the University of Konstanz giving talks on their research into autonomous robot navigation, neural mechanisms of compass orientation in insects and computational approaches to motor control.

The nuclear envelope is a lipid bilayer that encases the genome and provides a physical boundary between the cytoplasm and the nucleoplasm. While the nucleus is typically depicted as a sphere encircled by a smooth surface of nuclear envelope, the smooth exterior can be interrupted by tubular invaginations of the nuclear envelope into the deep nuclear interior. Such structures are termed the "nucleoplasmic reticulum." Increased frequency of nuclear envelope invagination occurs in disease states including various cancers, viral infections, and laminopathies, a group of heterogeneous disorders that arise due to mutations in the gene encoding lamin A. A significant increase in the frequency of nuclear envelope invaginations in the human Alzheimer's disease brain has recently been reported. Nuclear envelope invaginations are caused by pathogenic tau, one of the two major pathological hallmarks of Alzheimer's disease. Pathogenic tau-induced dysfunction of the lamin nucleoskeleton drives nuclear envelope invagination and consequent accumulation of polyadenylated RNA within invaginations, both of which drive neuronal death. Our ongoing studies suggest that maintaining proper cytoskeletal, nucleoskeletal, and genomic architecture are critical for survival and function of adult neurons.

Despite the importance of early diagnosis of dementia for prognosis and personalised interventions, we still lack robust tools for predicting individual progression to dementia. We propose a trajectory modelling approach that mines multimodal data from patients at early dementia stages to derive individualised prognostic scores of cognitive decline Our approach has potential to facilitate effective stratification of individuals based on prognostic disease trajectories, reducing patient misclassification with important implications for clinical practice.

Neurodegenerative diseases such as Alzheimer’s or Parkinson’s are devastating conditions with poorly understood mechanisms and no cure. Yet, a striking feature of these conditions is the characteristic pattern of invasion throughout the brain, leading to well-codified disease stages associated with various cognitive deficits and pathologies. How can we use mathematical modelling to gain insight into this process and, doing so, gain understanding about how the brain works? In this talk, I will show that by linking new mathematical theories to recent progress in imaging, we can unravel some of the universal features associated with dementia and, more generally, brain functions.

We seek to understand the control of behavior – by animals, their brains, and their neurons. Reiser and his team are focused on the fly visual system, using modern methods from the Drosophila toolkit to understand how visual pathways are involved in specific behaviors. Due to the recent connectomics explosion, they now study the brain-wide networks organizing visual information for behavior control. The team combines explorations of visually guided behaviors with functional investigations of specific cell types throughout the fly brain. The Reiser lab actively develops and disseminates new methods and instruments enabling increasingly precise quantification of animal behavior.

The main transmission mode of the COVID-19 disease is through virus-laden aerosols and droplets generated by expiratory events, such as breathing and sneezing. Patients with respiratory diseases are typically treated with oxygenation devices in hospitals, homes, and other settings where they increase the risk of spreading the disease to caregivers and first responders. Here, I will discuss a systematic study of aerosol and droplet dispersal through the air and their final deposition on surfaces. Through laser and fluorescent imaging techniques, we measure the volumetric spatial-temporal dynamics of droplet dispersal while varying rheological properties of the mucosaliva. We then demonstrate that a standard nose and mouth mask reduces the amount of mucosaliva dispersed by a factor of at least a hundred. Our ongoing collaborations with doctors and respiratory therapists from the Baystate Medical Hospital are developing new guidelines to help mitigate disease spread in a hospital setting.

Inflammation is a key component of the innate immune response. Primarily designed to remove noxious agents and limit their detrimental effects, the prolonged and/or inappropriately scaled innate immune response may be detrimental to the host and lead to a chronic disease. Indeed, there is increasing evidence suggesting that a chronic deregulation of immunity may represent one of the key elements in the pathobiology of many brain disorders. Microglia are the principal immune cells of the brain. The consensus today is that once activated microglia/macrophages can acquire a wide repertoire of profiles ranging from the classical pro-inflammatory to alternative and protective phenotypes. Recently, we described a novel ribosome-based regulatory mechanism/checkpoint that controls innate immune gene translation and microglial activation involving RNA binding protein SRSF3. Here we will discuss the implications of SRSF3 and other endogenous immune regulators in deregulation of immunity observed in different models of brain pathologies. Furthermore, we will discuss whether targeting SRSF3 and mRNA translation may open novel avenues for therapeutic modulation of immune response in the brain.

Three complementary lectures explain the development and applications of optogenetics. Gero Miesenböck traces optical control of neuronal activity from early experiments to studies of sleep, including neurons that respond to accumulated sleep need and molecular mechanisms governing sleep pressure. Peter Hegemann connects microbial photobiology to channelrhodopsins that control neuronal membrane voltage. He discusses their use in studying development and learning, and prospects for extending optical control to enzymes, transcription and translation. Georg Nagel follows the discovery and expression of microbial rhodopsins with light-sensitive transport and enzymatic functions. His examples include Channelrhodopsin-2 and its H134R variant, halorhodopsin, the more light-sensitive ChR2/XXL, the rhodopsin guanylyl cyclase Cyclop, and anion-channelrhodopsins in tobacco plants. Together, the talks connect molecular mechanisms with experimental control of neural and plant systems.

The combination of two-photon microscopy recordings and powerful calcium-dependent fluorescent sensors enables simultaneous recording of unprecedentedly large populations of neurons. While these sensors have matured over several generations of development, computational methods to process their fluorescence are often inefficient and the results hard to interpret. Here we introduce Suite2p: a fast, accurate, parameter-free and complete pipeline that registers raw movies, detects active and/or inactive cells (using Cellpose), extracts their calcium traces and infers their spike times. Suite2p runs faster than real time on standard workstations and outperforms state-of-the-art methods on newly developed ground-truth benchmarks for motion correction and cell detection.

The Stabilized Supralinear Network is a model of recurrently connected excitatory (E) and inhibitory (I) neurons with a supralinear input-output relation. It can explain cortical computations such as response normalization and inhibitory stabilization. However, the network's connectivity is designed by hand, based on experimental measurements. How the recurrent synaptic weights can be learned from the sensory input statistics in a biologically plausible way is unknown. Earlier theoretical work on plasticity focused on single neurons and the balance of excitation and inhibition but did not consider the simultaneous plasticity of recurrent synapses and the formation of receptive fields. Here we present a recurrent E-I network model where all synaptic connections are simultaneously plastic, and E neurons self-stabilize by recruiting co-tuned inhibition. Motivated by experimental results, we employ a local Hebbian plasticity rule with multiplicative normalization for E and I synapses. We develop a theoretical framework that explains how plasticity enables inhibition balanced excitatory receptive fields that match experimental results. We show analytically that sufficiently strong inhibition allows neurons' receptive fields to decorrelate and distribute themselves across the stimulus space. For strong recurrent excitation, the network becomes stabilized by inhibition, which prevents unconstrained self-excitation. In this regime, external inputs integrate sublinearly. As in the Stabilized Supralinear Network, this results in response normalization and winner-takes-all dynamics: when two competing stimuli are presented, the network response is dominated by the stronger stimulus while the weaker stimulus is suppressed. In summary, we present a biologically plausible theoretical framework to model plasticity in fully plastic recurrent E-I networks. While the connectivity is derived from the sensory input statistics, the circuit performs meaningful computations. Our work provides a mathematical framework of plasticity in recurrent networks, which has previously only been studied numerically and can serve as the basis for a new generation of brain-inspired unsupervised machine learning algorithms.