Animal model: Drosophila melanogaster.
Methods: volume EM (hemibrain connectome - https://elifesciences.org/articles/57443-, and FAFB dataset - https://www.sciencedirect.com/science/article/pii/S0092867418307876?via%3Dihub-), NBLAST (https://www.sciencedirect.com/science/article/pii/S0092867418307876?via%3Dihub), python and R code.
Main findings: The Authors perform an extensive data analysis and many are the important findings. Amongst those it’s worth to mention:

1. Obtaining a system wide understanding of the olfactory information flow by using LAYERS of neurons. Neuronal layers (mean path length from the sensory periphery to any neuron) are more complex and informative than neuronal orders (shortest path length), but still human intelligible.
2. Quantitative description of inputs and outputs of olfactory neurons up to third order.
3. By comparing the two datasets, the Authors were able to investigate stereotypy and lateralization.
4. High level of stereotypy across animals: in most of the cases it was possible to identify the same neurons (as for neuronal type) in different animals. More differences are found in larval born secondary neurons than in primary born embryonic neurons.
5. Lateralization to respond to lateralized odor sources.
6. Different pathways for odor and thermal/hygrosensory information.

Conclusions: I had the pleasure to present this important paper at our group JC with my colleague Kara Fulton. The paper itself is a massive work with a huge amount of data. Moreover, it is well presented and easy to read (despite some obscure sentences). In my opinion, it is a fundamental piece of work not only for the drosophila community, for the presented findings and the tools developed to analyze the data. It must be noted that the Authors made a big effort in combining morphological data with pre-existing data (functional and molecular).
Further questions: Would the cluster/ classification change if molecular data would be integrated? How ‘useful’ would be this dataset without the pre-existing literature? How much the stereotypy would be preserved after learning? Given the higher number of neurons in vertebrate brains (fish and mouse), how much stereotypy would be conserved?