Systematic outcome scoring for enhancing in vivo neurotherapeutic testing applied to perinatal brain injury

Official code for the analyses performed in the paper "Systematic outcome scoring for enhancing in vivo neurotherapeutic testing applied to perinatal brain injury". This code is under a license approved by the open source initiative.

1. System requirements

OS: Linux Debian Dependencies: file requirements.txt

2. Citation

@Unpublished{bokobza2025systematic,
  title={Systematic outcome scoring for enhancing in vivo neurotherapeutic testing applied to perinatal brain injury},
  author={Bokobza, Cindy and Réda, Clémence and Nair, Syam et al.},
  note={Under review},
  year={2025},
}

3. Installation guide (estimated install time: ~1min)

It is strongly advised to install the Conda tool to create a virtual environment, and Pip for installing dependencies:

conda create --name scoring python=3.6 -y
conda activate scoring
python3 -m pip install -r requirements.txt

4. Demo (estimated runtime: <2h)

4.a. Download the data

Download the data as described in the paper. The raw and processed data will be available on GEO at manuscript acceptance.

Filename	Administration protocol (time and mode)
P7 (inas P5)	P7 / INAS / P5
results_user_run409	P7 / IV / P5
results_user_run412	P12 / INAS / P10
results_user_run413	P12 / IV / P10
results_user_run415	P22 / IV / P20
results_user_run416	P22 / INAS / P20

4.b. Compute Characteristic Direction (CD) signatures and the cosine scores

conda activate scoring
## Apply the pipeline described in the paper for each batch
python3 -m ranking_conditions "P7 (inas P5)"
python3 -m ranking_conditions "results_user_run409"
python3 -m ranking_conditions "results_user_run412"
python3 -m ranking_conditions "results_user_run413"
python3 -m ranking_conditions "results_user_run415"
python3 -m ranking_conditions "results_user_run416"

After launching the command

python3 -m ranking_conditions "<batch_name>"

a folder named "<batch_name>" will be created, containing (1) a file named "CD_signatures.csv" with the signatures for each condition in the batch (PBS, Dose (low, medium, high)), (2) a file named "dose_ranking.csv", (3) 4 individual files containing each signature (PBS, Dose1, Dose2, Dose3). The concatenation of each "dose_ranking.csv" gives the ranking table with cosine scores.

4.c. Compare the ranking scores (cosine similarity) with the size of the gene support

The support for a cosine score on two signatures is the set of genes which are present in both signatures.

conda activate scoring
python3 -m correlation_score_geneset

This command returns a figure "correlation_score_intersections.png" plotting cosine scores against the size of the gene support, and computes the R2 metric between those two measures (R2=0,13) and the Pearson's R (r=-0.36, p-value=0.14).

4.d. Compare CD signatures with DESeq-based signatures (histogram)

conda activate scoring
python3 -m compare_analyses "" CD ## plot figures for CD signatures
python3 -m compare_analyses "" DESeq2 ## plot figures for DESeq2 signatures