Our Vision

A few thoughts on what motivates our lab’s efforts in computational neuroscience. A few words on a certain perceived deficiency in contemporary neuroscience and on an approach to neuroscience that is better suited to the Indian context.

Lack of deep and universal brain theories:
Contemporary neuroscience seems to suffer from a deep dichotomy that has its roots in a lopsided development of the field. On one hand, an ocean of brain data is available and is being generated every day, while progress in creation of deep and encompassing theories that make sense of such data is lagging behind. Large scale brain projects, which mark the trend of the day, are predominantly preoccupied with data generation, unprecedented in scale and reach, without a commensurate emphasis on creation of deep neural theories. Current neuroscience seems to be at a stage of development that is comparable to the astronomy of Ptolemy, or much later that of Tycho Brahe, when large amounts of star data was collected without any insights into the great principles or patterns that underlie such data.

The malaise of reductionism:
A certain pervasive reductionism is seen in most computational approaches to understand brain function. There is a tendency to explain all levels of brain’s functional hierarchy in terms of the lowest level – the level of cellular and molecular neuroscience. But such an approach is untenable and unrealistic when we compare neural theories with theorizing in physics and engineering. Physics offers us a range of theories, formulated at different levels of abstraction, each suitable to describe a given aspect of the physical world. Physical theories would fail miserably, and engineering would be impossible, had there been an absurd insistence on explaining all physical phenomena in atomic terms. (Imagine a study of airflow around an aircraft wing using molecular dynamics simulations!) Therefore a lot of computational neural models are data fitting models, like the Ptolemaic epicycles in astronomy, and do not possess the depth and versatility of a universal law like the inverse square law of gravitation, or the laws of electromagnetism. A veritable Principia of neuroscience is the need of the hour.

Deficiencies in applied neuroscience:
The ill effect of the absence of deep universal brain theories is seen most significantly in applied neuroscience of which there are two broad domains: clinical neuroscience and neurotechnology.  Since our understanding of the brain at large scale is still quite poor, the poverty haunts our understanding of neurological and neuropsychiatric disorders. We rarely have a rigorous, mechanistic understanding of how causes at cell and molecular level are manifest as symptoms of the disease at behavioral or whole body level. This lacuna is particularly striking in our understanding of complex neuropsychiatric diseases like Schizophrenia or neurodegenerative diseases like the Parkinson’s disease. Therefore, what is actually meted out to the patient as treatment – whether in terms of choice of drugs, or deep brain stimulation protocols - is a mass of empirical procedures and glorified thumb rules.

Therefore there is a profound need to launch a FOCUSED SEARCH FOR DEEP, INTEGRATIVE THEORIES OF THE BRAIN and improve our understanding of the brain at large scale or multi-scale level. These theories must explain how different levels of brain (molecular, cellular, local circuit, nucleus, subsystem…) work together and how causes at one level are manifest as effects at another (e.g. does a given Parkinson’s drug cure gait abnormalities?). 

Interdisciplinary teams dedicated to Neural disorders:

We suggest the following approach to realize the above objective. Computational model plays a central role in this approach.
Computational modeling research must be treated, not as an independent research methodology, but as a base for integration, as a glue that can paste together data from the staggering range of existing experimental methods of neuroscience. Small dedicated interdisciplinary teams of neuroscientists must be created. A typical team of this kind will have a computational modeler, an imaging expert, an electrophysiologist, a molecular biologist, a cell biologist,  a clinician, a biomedical engineer and a pharmacologist. The experimentalists probe the disease using very specific paradigms drawn from their respective research domains. In such a focused effort, the experimentalists do not have the luxury endlessly characterizing the system at hand in explosive detail. The work of these teams is like solving a jigsaw puzzle. Each team member must completely and exclusively dedicate himself/herself in filling a missing piece in that puzzle.  The results of their findings are integrated by the computational modeler who aims to create a comprehensive, computational theory of the respective neurological disease.
The theory developed will have a many-sided, positive impact. In pharmacology, the theory will have impact on drug design, drug delivery, drug mixture selection, dose determination etc. In deep brain stimulation, the theory will determine the site of stimulation and stimulation protocols. Similar impact can be anticipated in development of cognitive and physiotherapies for specific neurological/neuropsychiatric diseases.

Relevance to India:
Understanding the mechanistic basis of intelligence and creating an artificially intelligent organism is often touted as the holy grail of modern brain research. Such an ambitious aim might be more suitable for the developed world. But considering the socioeconomic conditions of a large country like India, we believe that a more appropriate national objective must be in the area of TRANSLATIONAL NEUROSCIENCE. Such an objective could ideally pursued by launching specific missions in various neurological and neuropsychiatric disorders, and by creating interdisciplinary teams of the kind outlined above. If such focused, interdisciplinary teams succeed, it might lead to a tremendous revolution in the field of clinical neuroscience. It will lift clinical neuroscience, that is currently immersed in dominantly empirical and experience-driven traditions, to a level of a solid scientific practice that rests on deep theoretical foundations.

V. Srinivasa Chakravarthy
CNS Team,
Department of Biotechnology, IITM.