The fly connectome reveals a path to the effectome.

Area of Science:

• Neuroscience
• Computational Neuroscience
• Systems Neuroscience

Background:

• Understanding the nervous system requires a causal model, but existing connectomes lack information on the strength of neural connections.
• The whole-brain fly connectome details synaptic paths but not the in vivo effect strength between neurons.

Purpose of the Study:

• To develop an efficient strategy for learning a causal model of the fly brain, termed the 'effectome'.
• To estimate causal effects and improve model efficiency using stochastic optogenetic perturbation data and the connectome as a prior.

Main Methods:

• Introduced a combined experimental and statistical strategy for causal modeling.
• Proposed an estimator for a linear dynamical model using optogenetic perturbation data.
• Utilized the fly connectome as a prior to enhance estimation efficiency.

Main Results:

• Validated the estimator in simulations, showing it approximates nonlinear dynamics.
• Identified dominant circuits within the fly nervous system, comprising small neuron populations.
• Re-discovered known circuits and generated testable hypotheses about neural dynamics.

Conclusions:

• Fly whole-brain dynamics are largely driven by independent, small-scale circuits.
• Neuron-level imaging, stimulation, and identification are feasible for studying these circuits.
• A causal model of the fly brain is achievable with this approach.