top of page
Single cell transcriptomics and lineage tracing on Parkinson's disease post-mortem data
In this project we studied the gene expression changes associated with idiopathic Parkinson's disease. We apply non-linear dimensionality reduction, variational auto-encoders, diffusion maps and computer simulations on single cell RNA-seq data derived from patients with Parkinson's disease. It allows to track the cell type specific gene expression changes associated with Parkinson's disease. We identified that there are common groups of genes associated with disease development identified by trajectory inference and gene network correlation analysis (WGCNA).
AD/PD 2022
Transcriptomic cell type analysis of the Alzheimer's brain
In this project we study the properties of human cortical neurons using 10x RNA-seq in the context of Alzheimer disease. We develop the reference dataset for Middle Temporal Gyrus using the novel clustering pipeline. Then we train classifiers based on the reference dataset to identify known neurons in the Alzheimer tissue. It allows us to investigate the cell type proportion changes and RNA trajectories associated with the Alzheimer disease. We do this work in close collaboration between the Allen Institute, University of Washington and Kaiser permanente.
Society for Neurosciences 2021
Analysis of human cortical neurons using patch-seq
In this project we study the properties of human cortical neurons derived from epilepsy and tumor patients. We link the single neuron electrophysiological properties to various brain pathologies and found neurons that are not affected. This allows to investigate the similarities between mouse and human neurons, providing the insight into the mechanisms of evolution of the neocortex. We do this work in collaboration with BICCN consortium directed by Ed Lein at Allen Institute.
Society for Neurosciences 2019
Multi-modal analysis of human epilepsy in hippocampus
In this work we aim to study the electrophysiological and morphological properties of single neurons derived from human epileptic tissue. We analyze the data that comes from patients with temporal lobe epilepsy. Specifically we look into pathological morphological and electrophysiological changes on single neuron level. To capture the single neuron properties we generate the detailed biophysical models of neurons and networks to test various scenario of seizure generation in silico. On the slides is present the response of the model cell to various current injections, experiment is in red, model is in blue. In this project we actively collaborate with Swedish medical center, working together with Costas Anastassiou and Jonathan Ting.
Society for Neurosciences 2018
K/Cl Homeostasis in human epilepsy in hippocampus
We developed the model of seizure initiation in temporal lobe epilepsy. Specifically we focused on the role of KCC2 contransporter which is responsible for mainaiting the baseline extracellular potassium and intracellular chloride levels in neurons. Recent experimental data has shown that this molecule is absent in the significant group of pyramidal cells in epileptic patients which suggest its epileptogenic role. In this project we looked into the consequences of this pathology from single neuron and neural network perspective. On the picture you could see the response of the network when there is 30% of KCC2-deficient pyramidal cells. We carried out this project in close collaboration with Richard Miles group at the Salpetriere Hospital in Paris.
Society for Neuroscience 2016
Inverse stochastic resonance in Purkinje Cells
We studied the role of synaptic noise in Purkinje cells. We investigated the effect of spike inhibition caused by noise current injection, so-called Inverse Stochastic Resonance (ISR). This effect has been previously found in single neuron models while we provided the first experimental evidence. On the slides you could see the model simulations that reproduce the experimentally observed ISR for particular variance of the noise input. As you could see on the picture different amplitudes of noise could significantly change the spiking behavior. We used methods of information theory and dynamical systems to study this phenomenon. In this project we collaborated with Michael Hausser's Neural Computation Laboratory at the University Colledge London.
Society for Neuroscience 2015
bottom of page