Skip to main content

Design Principles in Encoding Complex Noisy Environments

Final Report Summary - ENCODE (Design Principles in Encoding Complex Noisy Environments)

In this project we revealed design principles by which animals encode perplexing environments, and then integrated and propagated the information in the neural network. these principles include:
1. When having a limited compact sensory system, animals use sparse coding to efficiently code the environemnt.
2. When navigating in chemical gradients, animals developed a sophiticated mechanism to compute chemical derivatives to reach quickly reach the target.
3. Moreover, we revealed that animals typically make rational decisions when required to choose between various options. However, in sometimes they behave irrationally, and we extracted the nerual basis that underlies this irrational behavior.

To address these questions we developed novel experimental systems: (1) A multi-Animal Tracker that allows high-throughput screening of
interesting conditions by tracking multiple animals simultaneously. Moreover, fine features of locomotion and navigation in chemotaxis gradients towards food cues can be extracted. (2) A Tracker that measures neural activity in freely-behaving animals. This tracker allows correlating between the neural activity and computation done in circuits to behavioral outputs such as chemotaxis and decision making. (3) In addition, to focus on target interesting circuits we generated transgenic lines which express a combination of Channerhodopsin, Archaerhodopsin and GCaMP to activate, inhibit and measure from individual target neurons, respectively.