Structural and genetic determinants of zebrafish functional brain networks

Code associated with the "Structural and genetic determinants of zebrafish functional brain networks" paper, currently available on Science Advances, 11(28), eadv7576.

Repository overview

All data analyses from the paper are conducted in multiple Jupyter notebooks that are mostly split into 2 categories: Some that compute results for specific figures of the paper (FigureX-Analysis notebooks), and some that take these results to generate figures (FigureX-Layout notebooks) for subsequent modifications using Inkscape. Notebooks import functions in adjacent .py files and perform analyses on a processed calcium imaging dataset that can be downloaded here (roughly 20 GB of data). If notebooks are executed in the correct order, all figures can be reproduced, with the exception of a few panels that require raw data (over a terabyte, available upon reasonable request). Supplementary analyses are done in Supp- notebooks.

Notebooks list

• Figure1-Analysis : This notebook briefly inspects the calcium imaging dataset, counts neurons in every brain region, excludes regions with few cells, detrends fluorescence time series, computes functional networks, and performs the inter-individual similarity/fingerprinting analysis; statistical validations of fingerprinting are conducted in a separate notebook.
• Figure1-Layout : Loads previously calculated results and generates Figure 1.
• Figure2-Analysis : Uses SC and FC matrices to evaluate the structure-function relationship of zebrafish brain networks.
• Figure2-Layout : Loads previously calculated results and generates Figure 2.
• Figure3-Layout : Calculates SC and FC communities, then evaluates their modularity, overlap, and spatial extent, relative to multiple null models.
• Figure3-Layout : Loads previously calculated results and generates Figure 3.
• Figure4-Analysis : Identifies high-amplitude brain activity patterns and evaluates their modularity; also projects raw calcium volumes into the brain atlas.
• Figure4-Detrending : Loops over the raw calcium data to compute a pixel-wise linear regression and generate detrended calcium videos.
• Figure4-Layout : Loads previously calculated results and generates Figure 4.
• Figure5-Analysis : Identifies motor and visual neurons using regression analysis, then visualizes these neurons in the brain atlas.
• Figure5-Layout : Loads previously calculated results and generates Figure 5.
• Figure6-Analysis : Computes SC and FC gradients, compares them, evaluates their statistical significance, then computes the sensorimotor index and correlates it with the main FC gradient.
• Figure6-Layout : Loads previously calculated results and generates Figure 6.
• Figure7-Analysis : Computes regional gene expression levels, generates surrogates, conducts the simulated annealing optimization, then generates projections of the identified genes.
• Figure7-Layout : Loads previously calculated results and generates Figure 7.
• Supp-Figures : Generates Supplementary Figures S1, S2, S5, S8, S13, S14, S15, S16, S17, S19, S20, S21, S22, S23, S25.
• Supp-Averaging : Evaluates the effect of averaging signals to compute FC, related to Supplementary Figure S3.
• Supp-DatasetsComparison : Computes FC matrices from different calcium imaging datasets and compares them, related to Supplementary Figure S4.
• Supp-Fingerprinting : Compares different methods to fingerprint the identity of larvae from calcium imaging data, then generates Supplementary Figure S6.
• Supp-FC-vs-Signal : Evaluates the correlation between FC and signal intensity levels, related to Supplementary Figure S7.
• Supp-ComputingSC : Uses single-neuron reconstructions to compute SC matrices, related to Supplementary Figure S9.
• Supp-SC-Validations : Validation of the boundary expansion procedure for SC calculation, related to Supplementary Figure S9.
• Supp-SCCM : Defines the connectivity null model, generates null connectivity matrices from SC matrices generated beforehand, and generates Supplementary Figures S10, S11, & S12.
• Supp-SparseSC : Replication of key results using the sparse SC matrix, related to Supplementary Figure S18.
• Supp-Gradients : Computes FC gradients in various FC definitions, then generates Supplementary Figure S24.

Installation

Download this repository, then run the notebooks in a Python environment equipped with the appropriate packages, which can all be pip installed easily. Some notebooks will output files in a Results/ folder, which should be manually created and placed adjacent to the notebooks.

Citation

If you use this repository for your own work, please cite the paper:

Légaré, A., Lemieux, M., Boily, V., Poulin, S., Légaré, A., Desrosiers, P., & De Koninck, P. (2025). Structural and genetic determinants of zebrafish functional brain networks. Science Advances, 11(28), eadv7576. https://doi.org/10.1126/sciadv.adv7576

BibTeX:

@article{legare2025structural,
  title = {Structural and genetic determinants of zebrafish functional brain networks},
  author = {L{\'e}gar{\'e}, Antoine and Lemieux, Mado and Boily, Vincent and Poulin, Sandrine and L{\'e}gar{\'e}, Arthur and Desrosiers, Patrick and De Koninck, Paul},
  journal = {Science Advances},
  volume = {11},
  number = {28},
  pages = {eadv7576},
  year = {2025},
  doi = {10.1126/sciadv.adv7576},
  url = {https://doi.org/10.1126/sciadv.adv7576}
}

Authors

• Antoine Légaré (antoine.legare.1@ulaval.ca)
• Patrick Desrosiers (patrick.desrosiers@phy.ulaval.ca)

For any questions regarding the repository, please contact us!