Unpacking the Biology of Psychiatric Genetics using Cell Painting

Here, we propose to adapt the Cell Painting assay to interrogate traits in neuronal cells, and apply it to a cohort of 48 cell lines carrying the 22q11.2 deletion, a genetic variant strongly associated with psychiatric disease, to test the utility of neuronal Cell Painting in identifying disease relevant phenotypes in a high-throughput setting. Our work will bring together groups at the Imaging Platform and the Stanley Center to create a new strategy that could be integrated into multiple existing platforms at the Broad to enable, for the first time, the scaled investigation of neuronal profiles. We anticipate that the workflows we will create will facilitate phenotypic screening of neurons at a scale that begins to match the transcriptional revolution, and constitute a key technology to help move from genetics to cellular phenotypes to actionable biology and mechanisms. Gaining insight into neuronal morphology in health and disease will illuminate previously unknown aspects of neuronal biology, enable the interrogation of the effect of the hundreds of genetic risk variants on cellular phenotypes, and greatly complement several existing technologies pioneered at the Broad Institute such as CRISPR screens, drug screens, optical profiling and in situ sequencing to catalyze unprecedented discoveries that link genes and perturbations to neuronal phenotypes.

Dataset summary

There are 48 IPSC lines available for this project from the 22q cohort:

| Metadatalinesource | Metadatalinecondition | n | |:---------------------|:------------------------|---:| | human | control | 22 | | human | deletion | 22 | | isogeniccontrol | control | 2 | | isogenicdeletion | deletion | 2 |

TODO: All this information should be moved to the Project Profiler airtable.

TODO: Fill in information about number of features

| Experiment | Plate | Features | Magnification | Profiles | Notes | |------------------------------|------------------|-------------------------------------------------------|---------------|-------------------------------------------------------------------------------------------------------------------------------------------------------|-----------------------------------------------------------------------------------------------------------------------------------------------------| | NCP Stem 1 | BR_NCP_STEM_1 | Cell Painting (n=4293) | 20x | GitHub | | | NCP Progenitor 1 | BR00127194 | Cell Painting (n=4295, includes 4 branching features) | 20x | GitHub | This is a repeat of an experiment that failed (notes). | | NCP Neuron 1 - Cell Painting | BR00132672 | | 20x | S3 | (notes) | | NCP Neuron 1 - Cell Painting | BR00132672 | | 63x | S3 | same ^^^ | | NCP Neuron 1 - Cell Painting | BR00132673 | | 20x | S3 | same ^^^ | | NCP Neuron 1 - Cell Painting | BR00132673 | | 63x | S3 | same ^^^ |

Failed experiments

| Experiment | Plate | Features | Profiles | Notes | |------------------|-------------------------|------------------------|-------------------------------------------------------------------------------------|-----------------------------------------------------------------------------------------------------------------------------------------------| | NCP Progenitor 1 | BR_NCP_PROGENITORS_1 | Cell Painting (n=4293) | GitHub | This was the first attempt but it failed (notes) | | | BR_NCP_PROGENITORS_1 | Branching (n=23) | GitHub | Same as above, only branching metrics |

Profiles from newer datasets (2022 onwards) are in this data repo https://github.com/broadinstitute/2019_05_28_Neuronal_Cell_Painting

We have RNA-Seq data (Nehme, Pietiläinen, et al., submitted) for 20 healthy controls and 28 patients with 22q deletion, across 3 stages:

• D0 (undifferentiated stem cells)
• D4 (progenitors, with GFP)
• D28 (neurons)

Computational environment

Python environment

We use mamba to manage the computational environment.

To install mamba see instructions.

After installing mamba, execute the following to install and navigate to the environment:

```bash

First, install the conda environment

mamba env create --force --file environment.yml

If you had already installed this environment and now want to update it

mamba env update --file environment.yml --prune

Then, activate the environment and you're all set!

environmentname=$(grep "name:" environment.yml | awk '{print $2}') mamba activate $environmentname ```

R

We use renv to reproduce R code. We recommend using RStudio as your IDE.

Checkout this repository and then load the project neuronal-cell-painting.Rproj in RStudio. You should see this

```text

Bootstrapping renv 0.13.1 --------------------------------------------------

• Downloading renv 0.13.1 ... OK
• Installing renv 0.13.1 ... Done!
• Successfully installed and loaded renv 0.13.1.
• Project '~/Downloads/neuronal-cell-painting.Rproj' loaded. [renv 0.13.1]
• The project library is out of sync with the lockfile.
• Use renv::restore() to install packages recorded in the lockfile. ```

Now run renv::restore() and you're ready to run the R scripts in this repo.

Note: If you end up with issues with compiling libraries and you are on OSX, it's probably something to do with the macOS toolchain for versions of R starting at 4.y.z. being broken. Follow these instructions to get set up.

Creating a new R notebook