Projects

Current Projects

Applied research

Therapeutic test-bench for PD (TTB)

AIM - Developing a therapeutic test-bench for Parkinson's disease (HTM)

Objectives

Modeling the neuropharmacology aspects of drug action
- Pharmacokinetics
- Pharmacodynamics
Modeling chronic levodopa treatment consequences and motor dysfunctions in Parkinson's disease
- Wearing-off phenomenon (predictable motor fluctuations)
- On-off phenomenon (unpredictable motor fluctuations)
- Levodopa-induced dyskinesias.
Modeling the impact of other drugs in Parkinson's disease
- Dopamine Agonists
- MAO Inhibitors
- COMT inhibitors
- Cholinergic Drugs
- Serotonergic Drugs

References: -

Chakravarthy VS, Moustafa AA (2018) Computational Neuroscience Models of the Basal Ganglia. Singapore: Springer Singapore. Available at: http://dx.doi.org/10.1002/1531-8257(200009)15:5%3C762::AID-MDS1002%3E3.0.CO%5Cnhttp://www.ncbi.nlm.nih.gov/pubmed/11009178.Muralidharan V, Mandali A, Balasubramani PP, Mehta H, Chakravarthy VS (2016) A scalable cortico‑basal ganglia model to understand the neural dynamics of targeted reaching. BMC Neurosci 17:2016.

Simplified Model of Substantia Nigra pars compacta (SNc)

AIM - Developing a simplified model of SNc (both soma and terminal)

Objectives

The biophysical model of SNc becomes too expensive in terms of computational complexity. In order to expand the pharmacological test-bench in terms of other adjunct drug administration, validating the side effects of drug interventions and modeling other therapeutic interventions a much more reduced model of SNc is important.
- Simplifying the SNc SOMA
- Simplifying the SNc Terminal
Methods
- Using spiking neural network such as Izhikevich neuronal model
- Representing input/output of SOMA in terms of current as input and firing as output
- Modelling terminals with firing as input and DA release as output
- Reduced model of drug administration
- Integration of simplified drug administration model with SNc terminal.

Developing a deep brain stimulation (DBS) test bench for PD

AIM - Pharmacological Interventions such as Levodopa medication helps to address the symptoms to an extent however with progression of the disease and prolonged period of dosage, the levodopa effect start to cease and patients start exhibiting side effects such as wearing off and dyskinesia's. DBS intervention helps to relieve symptoms to an extend and also reduces the amount of drug dosage required.

Objectives

To develop a spiky neuron model of SNc and STN
To simulate DBS effects in the model
Integrating the DBS test bench and the pharmacological test bench

Modelling Working Memory functions of the Basal Ganglia

AIM - To develop a model of Basal Ganglia using which we can test the working memory properties.

Objectives

Working memory refers to the ability to temporarily store and manipulate information necessary to execute complex cognitive tasks. In computational terms, working memory is thought to be encoded in terms of stable neural activation patterns, distinguishing it from longer term memories based on synaptic update.
Two brain structures are thought to be crucially involved in working memory – the prefrontal cortex and the basal ganglia – both of which receive projections from the midbrain dopaminergic nuclei.
In this study we focus on the working memory functions of the Basal Ganglia (BG).
Methods
- Modelling various tasks such as 1ax-2by, n-back and 2x5 tasks using which we can access the working memory properties

Modelling multiple tasks to replicate PD symptoms in order to formulate a comprehensive evaluation regime.

AIM - To model various array of tasks that can replicate PD impairments such as Gait, Eye movements, Handwriting, Speech, IGT, Working memory.

Objectives

To model wide array of tasks as mentioned above.
Find the meta parameters that represents dopamine, serotonine and norepinephrine .
Tune the meta parameters to replicate PD condition.
Model therapeutic strategies

Handwriting model using flip flop network

AIM - Handwriting analysis is one of the early stage detection mechanism of PD. In this work we develop a model that generates letters and model handwriting.

Objectives

To generate letters using Flip Flop network and temporal relationships in data.
To model the conditions such as progressive micrographia.
Use Handwariting information to predict the onset of PD.

Outcome - Currently we are able to satisfactority generate character sequences using a series of Flip Flop's using supervised learning technique.

References: -

Holla, P., Chakravarthy, S. (2016). Decision making with long delays using networks of ip-op neurons. 2016 International Joint Conference on Neural Networks (IJCNN), 2767-2773.
Garipelli Gangadhar, Denny Joseph, A.V. Srinivasan, Deepak Subramanian, Shiva Keshavan, V. Srinivasa Chakravarthy, A computational model of Parkinsonian handwriting that high lights the role of the indirect pathway in the basal ganglia. Human Movement Science, 2009 Oct;28(5):602-18
G. Gangadhar, D. Joseph, V.S. Chakravarthy, Understanding Parkinsonian Handwriting using a computational model of basal ganglia, Neural Computation, 20, 1–35 (2008)
Graves, Alex.(2013),"Generating sequences with recurrent neural networks." arXiv preprintarXiv:1308.0850.

Canonical Cortico-Basal Ganglia Model for Iowa Gambling Task (IGT)

AIM - To combine the decision making aspects to the assessment of motor performance using computational model .

Objectives

To develop an IGT gaming task and integrate to the model
To access the cognitive performance along with the motor performance
To test the impact of various therapeutic interventions
Data inversions using the results from clinical trials

Outside BG research

Glutamate-induced toxicity in CA1 (GIT-AD)

AIM - Modeling a spiking network model of excitotoxicity in CA1 during Alzheimer's disease (SEM-AD)

Objectives

Similar to PD, we hypothesis excitotoxicity in AD might also be precipitated by energy deficiency.
Propose novel therapeutic intervention for neuroprotection and also suggest disease-modifying strategies

References: -

Ong WY, Tanaka K, Dawe GS, Ittner LM, Farooqui AA (2013) Slow excitotoxicity in Alzheimer’s disease. J Alzheimer’s Dis 35:643–668.Nobili A et al. (2017) Dopamine neuronal loss contributes to memory and reward dysfunction in a model of Alzheimer’s disease. Nat Commun 8:14727 Available at: http://www.nature.com/doifinder/10.1038/ncomms14727

Completed Projects

Applied Research

A Generalized Reinforcement Learning-Based Deep Neural Network (GRL-DNN) Agent Model for Diverse Cognitive Constructs

AIM - .To develop a computational agent model of decision making which can explain important cognitive abilities including selective attention, response inhibition, distractor processing, and working memory properties.

Objectives

To simulate a computational model of various decision making tasks
The decision making tasks tests the selective attention, response inhibition, distractor processing and working memory processing abilities
To tune the performance of the cognitive tasks using meta parameters.
To build an inverse model where from the performance, the meta parameters can be predicted
In future, we look forward to model the disease conditions such as anxiety, hypertension, ADHD etc.

Outcome - To model diverse cognitive tasks for a healthy control subject.

References: -

Nair, S. S., Muddapu, V. R., Vigneswaran, C., Balasubramani, P. P., Ramanathan, D. S., Mishra, J., & Chakravarthy, V. S. (2022). ‘A Generalized Reinforcement Learning-Based Deep Neural Network (GRL-DNN) Agent Model for Diverse Cognitive Constructs. BioRxiv, 2022.06.17.496500. https://doi.org/10.1101/2022.06.17.496500

Integrated Multi-scale pharmacological model for Parkinson's Disease (MCBG)

AIM - To develop a multi scale computational model that can replicate the symptoms at the behavioural level by incorporating the key cellular and molecular mechanisms underlying PD pathology.

Objectives

To understand the impact of cell loss on motor performance
To simulate various cardinal symptoms of PD such as Bradykinesia, Tremor, Rigidity
To analyze the abnormal beta band oscillations caused due to the SNc cell loss.
To test the effect of LDOPA medication on reaching performance
To simulate the side effects such as wearing off and dyskinesias.
To optimize the drug dosage based on improving the performance and minimizing the side effects.

Outcome - Replication of PD symptoms using the model and testing the pharmacological intervention

References: -

Nair SS, Muddapu VR, Chakravarthy VS. A Multiscale, Systems-Level, Neuropharmacological Model of Cortico-Basal Ganglia System for Arm Reaching Under Normal, Parkinsonian, and Levodopa Medication Conditions. Front Comput Neurosci. 2022 Jan 3;15:756881. doi: 10.3389/fncom.2021.756881. PMID: 35046787; PMCID: PMC8762321.

Basic Research

Glutamate-induced toxicity in SNc (GIT)

AIM - Modeling a hybrid network model of excitotoxicity in SNc during Parkinson's disease (HEM)

Objectives

To decipher the mechanism behind excitotoxicity during energy deficiency in SNc
Propose novel therapeutic interventions for neuroprotection and also suggest disease-modifying strategies

Outcome - Spiking network model of excitotoxicity (SEM) and neuroprotective interventions (1,2)

References: -

Muddapu VR, Mandali A, Chakravarthy S V, Ramaswamy S (2018) A computational model of loss of dopaminergic cells in Parkinson’s disease due to glutamate-induced excitotoxicity. bioRxiv:1–69 Available at: https://www.biorxiv.org/content/early/2018/08/09/385138.Muddapu VR, Chakravarthy S V (2017) Programmed cell death in substantia nigra due to subthalamic nucleus-mediated excitotoxicity: a computational model of Parkinsonian neurodegeneration. In: BMC Neuroscience, pp 59 Available at: https://doi.org/10.1186/s12868-017-0371-2#Sec113.Rodriguez MC, Obeso JA, Olanow CW (1998) Subthalamic nucleus-mediated excitotoxicity in Parkinson’s disease: a target for neuroprotection. Ann Neurol 44:S175-88 Available at: http://www.ncbi.nlm.nih.gov/pubmed/9749591.Greene JG, Greenamyre JT (1996) Bioenergetics and glutamate excitotoxicity. Prog Neurobiol 48:613–634.

LDOPA-induced toxicity in SNc (LIT)

AIM - Modeling a hybrid network model of Levodopa-induced toxicity in SNc during Parkinson's disease (HLM)

Objectives

To decipher how exogenous LDOPA causing SNc neurodegeneration during energy deficiency
Propose novel therapeutic intervention for neuroprotection and also suggest disease-modifying strategies

References: -

Thornton E, Vink R (2015) Substance P and its tachykinin NK1 receptor: a novel neuroprotective target for Parkinson’s disease. Neural Regen Res 10:1403–1405 Available at: http://www.nrronline.org/text.asp?2015/10/9/1403/165505.Lipski J, Nistico R, Berretta N, Guatteo E, Bernardi G, Mercuri NB (2011) L-DOPA: A scapegoat for accelerated neurodegeneration in Parkinson’s disease? Prog Neurobiol 94:389–407 Available at: http://dx.doi.org/10.1016/j.pneurobio.2011.06.005.Fahn S (2005) Does levodopa slow or hasten the rate of progression of Parkinson’s disease? J Neurol 252:iv37-iv42 Available at: http://link.springer.com/10.1007/s00415-005-4008-5 .

Comprehensive SNc Model

AIM - To develop a detailed biophysical model that can explain the dynamics involved in different molecular processes of SNc such as mitochondria, glycolysis, ROS, Ldopa uptake mechanism. This is important in analyzing the extracellular dopamine released and how energy deficiency impacts the dopamine supply in the extra cellular space.

Objectives

How energy deficiency is impacting molecular processes

References: -

Vignayanandam R. Muddapu & V. Srinivasa Chakravarthy (2020) Influence of Energy Deficiency on the Molecular Processes of Substantia Nigra Pars Compacta Cell for Understanding Parkinsonian Neurodegeneration - A Comprehensive Biophysical Computational Model

Completed Projects from previous years

Modeling a hybrid network model of Levodopa-induced toxicity in SNc during Parkinson's disease (HLM) (Vignayanandam R. Muddapu)

To decipher how exogenous LDOPA causing SNc neurodegeneration during energy deficiency. Propose novel therapeutic intervention for neuroprotection and also suggest disease-modifying strategies

References: -

Influence of Energy Deficiency on the Molecular Processes of Substantia Nigra Pars Compacta Cell for Understanding Parkinsonian Neurodegeneration (Vignayanandam R. Muddapu)

Developed a detailed biophysical computational model that can explain the dynamics involved in different molecular processes of SNc such as mitochondria, glycolysis, ROS, Ldopa uptake mechanism. This is important in analyzing the extracellular dopamine released and how energy deficiency impacts the dopamine supply in the extra cellular space. The model also tries to analyze how energy deficiency is impacting molecular processes.

References: -

Modeling a hybrid network model of excitotoxicity in SNc during Parkinson's disease (HEM)

To decipher the mechanism behind excitotoxicity during energy deficiency in SNc. To Propose novel therapeutic interventions for neuroprotection and also suggest disease-modifying strategiesOutcome - Spiking network model of excitotoxicity (SEM) and neuroprotective interventions (1,2)References: -[1] Muddapu VR, Mandali A, Chakravarthy S V, Ramaswamy S (2018) A computational model of loss of dopaminergic cells in Parkinson’s disease due to glutamate-induced excitotoxicity. bioRxiv:1–69 Available at: https://www.biorxiv.org/content/early/2018/08/09/385138.[2] Muddapu VR, Chakravarthy S V (2017) Programmed cell death in substantia nigra due to subthalamic nucleus-mediated excitotoxicity: a computational model of Parkinsonian neurodegeneration. In: BMC Neuroscience, pp 59 Available at: https://doi.org/10.1186/s12868-017-0371-2#Sec113.Rodriguez MC, Obeso JA, Olanow CW (1998) Subthalamic nucleus-mediated excitotoxicity in Parkinson’s disease: a target for neuroprotection. Ann Neurol 44:S175-88 Available at: http://www.ncbi.nlm.nih.gov/pubmed/9749591.Greene JG, Greenamyre JT (1996) Bioenergetics and glutamate excitotoxicity. Prog Neurobiol 48:613–634.

A Neurocomputational Model of the Effect of Cognitive Load on Freezing of Gait in Parkinson's Disease ( Vignesh Muralidharan)

A computational model of the cognitive and motor cortico-basal ganglia loops that explains the effects of sensory and cognitive processes on FOG. The model simulates strong causative factors of FOG including decision conflict.

Probing the Role of Medication, DBS Electrode Position, and Antidromic Activation on Impulsivity Using a Computational Model of Basal Ganglia ( Alekhya Mandali )

Understanding the dynamics of impulsive behaviour in a tough decision making scenario for the people who undergoes a dopamine medication and DBS using a computational model.

A network model of basal ganglia for understanding the roles of dopamine and serotonin in reward-punishment-risk based decision making (Pragathi P. Balasubramani )

Understanding the risk based decision making and exploring the joint contribution of 5HT along with the DA in modulating the activities of the MSNs using a computational model.

Computational model of precision grip in Parkinson's disease: a utility based approach (Ankur Gupta)

A computational model that studies the variations in grip force in dopaine deficit conditions. The model focuses on BG action selection mechanism.

Modeling the Contributions of Basal Ganglia and Hippocampus to Spatial Navigation Using Reinforcement Learning (Deepika Sukumar)

A computational model that presents an integrated model for navigation where the contributions from Basal ganglia and Hipocampus are studied.

On the neural substrates for exploratory dynamics in basal ganglia: A model (Sanjeeva Kalva)

This computational study focuses on the exploratory dynamics of the STN-GPe basal ganglia subsystem. While the extreme values of daopaine sitches the system to Go/No-Go regimes, small intermediate values of dopamine supply enables the system to operate in an exploratory regime. This study explores the intermediate regime in detail.

Modelling Basal Ganglia for understanding Parkinsonian reaching movements: (K. N. Magdoom )

This computational models the arm reaching movement and studies the effect of PD in detail. The model is cast within the reinforcement learning (RL) framework with correspondence between RL components and neuroanatomy as follows: dopamine signal of substantia nigra pars compacta as the temporal difference error, striatum as the substrate for the critic, and the motor cortex as the actor. The indirect pathway consisting of the STN-GPe subsystem is the explorer.

A computational model of Parkinsonian handwriting that highlights the role of the indirect pathway in the basal ganglia (G.Gangadhar)

This computational model studies simulates the handwriting generation and analyses how the PD condition effect the sizes of written letters. The variation in handwriting is linked to the synchronous activity in the STN-GPe subsystem.

THE ROLE OF THE BASAL GANGLIA IN EXPLORATION IN A NEURAL MODEL BASED ON REINFORCEMENT LEARNING (D. SRIDHARAN)

This study models an Reinforcement learning based Basal Ganglia exploratory system. Here STN-GPe subsystem of BG is modelled as a pair of neuronal layers with oscillatory dynamics, exhibiting a variety of dynamic regimes such as chaos, traveling waves and clustering. Invoking the property of chaotic systems to explore state-space, this study suggest that the complex exploratory dynamics of STN-GPe system in conjunction with dopamine-based reward signaling from the Substantia Nigra pars compacta (SNc) present the two key ingredients of a reinforcement learning system.

Google Sites

Report abuse