Fellow's research: Variable wiring, consistent response - How do brains do it?

11 Mar 2020

Fellow's research: Variable wiring, consistent response - How do brains do it?

Nitin Gupta, Intermediate Fellow, IIT Kanpur

In our recent article in Nature Communications, we report the mechanisms that overcome the variability in wiring of brains in different individuals.

Most of us find the smell of an orange pleasing and the smell of a rotten fish stinking. This similarity in our perceptions is not because our brains are wired identically. In fact, it would be impossible to get identical brain wiring in different people. A human brain has about 100 billion neurons and over a 100 trillion connections – such a humongous amount of detail cannot be stuffed into our DNA, whose storage capacity is thousands of times smaller. Only the coarse layout of the brain circuits can be specified genetically, while the exact connections vary from individual to individual. (This is analogous to how the shapes of our fingers are quite similar, while the fingerprints vary from individual to individual). Given the variability in the brain wirings, how do most people manage to perceive the rotten fish as stinking?

We made use of the detailed experimental data available for the mushroom body, an area of the insect brain involved in processing smells. The neurons that provide input to the mushroom body are wired similarly across individuals. And so, their responses to smells are also stereotyped across individuals. But, the connections within the mushroom body vary from individual to individual. Thus, it was expected that the responses of all subsequent neurons would be highly variable.

We noticed that a previous research study had reported stereotyped responses in the output neurons of the mushroom body in flies. To confirm that this was not an anomaly, or something limited to flies, we checked the responses of the output neurons in locusts and found the same surprising result.

To figure out what was going on, we simulated these brain networks on a computer. We found that the output neurons get stereotyped responses because they combine the activities of a large number of internal neurons – the summed activity of many neurons becomes more consistent than the activity of a single internal neuron.

We identified other mechanisms that contribute to the stereotypic responses and supported the results of our simulations with mathematical modeling. We also came up with a new metric, called PRED, for measuring stereotypy in sensory responses. The research contributes to the understanding of how brains can respond reliably to important stimuli in the environment, despite the wiring variabilities.


Multiple network properties overcome random connectivity to enable stereotypic sensory responses. Aarush Mohit Mittal, Diksha Gupta, Amrita Singh, Andrew C. Lin, and Nitin Gupta. Nature Communications 11, no. 1 (2020): 1-15.

Banner image: A brain network in two different individuals (left and right graphs). Although the connections between the first two layers are different across the individuals, the response of the output neuron (blue) can be similar.

Watch this video to learn more about Nitin's work.