M1/M2 internship: Developing Genotype-Conditioned Artificial Drosophila Larvae Behavioural Data Through Bayesian Program Synthesis

Drosophila larvae have emerged as an ideal platform for simultaneously probing behaviour and the underlying neuronal computation^1–5. Modern genetic tools allow efficient activation or silencing of individual and small groups of neurons. Combining these techniques with standardised stimuli over thousands of individuals makes it possible to relate neurons to behaviour causally However, extracting these relationships from massive and noisy recordings requires the development of new statistically robust approaches.
Recently, in experimental settings to probe defensive actions⁶ or understand the implementation of neuromodulation in small neural neworks⁷, larva populations exhibited significant deviations from usual behavioural features. These deviations lead to either the redefinition of features usually associated with these behaviours or to challenges in detecting them.
We aim to develop a Bayesian Program Synthesis⁸ (BPS) methodology for producing synthetic data that mirrors key characteristics found in experimental recordings. This method will entail an inference phase to identify large-scale parameters characteristic of larva populations. It will incorporate various generative models to either replicate the behavioural traits associated with particular genotypes or create bespoke behavioural features. Additionally, we plan to use a collection of small generative programs to create larva data under specific action conditions. The efficacy of this approach will be demonstrated by applying our behaviour analysis pipeline to the synthetic data and assessing its ability to facilitate transfer learning with annotated behavioral datasets.

Fig 1) Example of artificial generated from a pre-defined larva genotype

References

1. Masson, J.-B. et al. Identifying neural substrates of competitive interactions and sequence transitions during mechanosensory responses in Drosophila. PLOS Genet. 16, e1008589 (2020).

2. Jovanic, T. et al. Competitive Disinhibition Mediates Behavioral Choice and Sequences in Drosophila. Cell 167, 858-870.e19 (2016).

3. Croteau-Chonka, E. C. et al. High-throughput automated methods for classical and operant conditioning of Drosophila larvae. eLife 11, e70015 (2022).

4. Vogelstein, J. T. et al. Discovery of Brainwide Neural-Behavioral Maps via Multiscale Unsupervised Structure Learning. Science 344, 386–392 (2014).

5. Winding, M. et al. The connectome of an insect brain. Science 379, eadd9330 (2023).

6. Lehman, M. et al. Neural circuits underlying context-dependent competition between defensive actions in Drosophila larva. 2023.12.24.573276 Preprint at https://doi.org/10.1101/2023.12.24.573276 (2023).

7. Tredern, E. de et al. Feeding-state dependent modulation of reciprocally interconnected inhibitory neurons biases sensorimotor decisions in Drosophila. 2023.12.26.573306 Preprint at https://doi.org/10.1101/2023.12.26.573306 (2023).

8. Lake, B. M., Salakhutdinov, R. & Tenenbaum, J. B. Human-level concept learning through probabilistic program induction. Science 350, 1332–1338 (2015).

Scientific or technical background required for work program

The successful intern should have one of the following backgrounds:

Statistical Physics, Applied Mathematics,
Statistics & Bayesian Inference

Some fluency in Python is expected.