BrainNetAnno 0.1.0

A toolkit for molecular annotation of brain networks, integrating transcriptome, neurotransmitters, and mitochondria.

BrainNetAnno

A toolkit for molecular annotation of brain networks, integrating transcriptome, neurotransmitters, and mitochondria.

Features

• CPU-only implementations (NumPy/SciPy/Scikit-learn/Statsmodels/Matplotlib)
• Shared utilities in BrainNetAnno.utils
• PLS pipelines: *_pls_cge.run_*_pls_pipeline
• Enrichment analyses: gene_celltype.run_pipeline, gene_layer.run_pipeline

Requirements

• Python >= 3.8
• Main dependencies (see pyproject.toml):
  • numpy, pandas, scipy, scikit-learn, statsmodels, matplotlib

Installation

From source:

pip install BrainNetAnno

Or build a wheel and install:

python -m pip install -U build
python -m build
pip install dist/BrainNetAnno-<version>-py3-none-any.whl

Quickstart (Python API)

• Transcriptome PLS-CGE pipeline:

from BrainNetAnno.transcriptome_pls_cge import run_transcriptome_pls_pipeline

run_transcriptome_pls_pipeline(
    t_values_file="path/to/fc_deviation.csv",          # square matrix CSV, no header
    gene_expression_file="path/to/gene_contrib.csv",   # contains 'Region_Pair'
    output_best_genes_csv="path/to/best_genes.csv",
    fig_outputfile="path/to/rmse_variance.tif",
)

• Cell-type enrichment:

from BrainNetAnno.gene_celltype import run_pipeline

res = run_pipeline(
    celltype_csv="path/to/celltypes_PSP.csv",      # columns: gene, class
    target_genes_csv="path/to/best_genes_SCZ.csv", # column: Gene Index
    output_csv="path/to/celltype_enrichment_results.csv",
)
print(res.head())

• Cortical layer enrichment:

from BrainNetAnno.gene_layer import run_pipeline

res = run_pipeline(
    layer_marker_path="path/to/41593_2020_787_MOESM3_ESM.xlsx",  # sheet: Table S4B
    target_genes_path="path/to/best_genes_all.csv",
    output_csv="path/to/ALL_gene_layer_analysis_results.csv",
)
print(res.head())

• Neurotransmitter PLS pipeline:

from BrainNetAnno.neurotransmitter_pls_cge import run_neurotransmitter_pls_pipeline

best_n, df = run_neurotransmitter_pls_pipeline(
    fc_matrix_path="path/to/fc_deviation.csv",      # square matrix CSV, no header
    nt_contrib_csv="path/to/nt_contrib.csv",        # contains 'Region_Pair'
    output_weights_csv="path/to/nt_weights.csv",
)
print(best_n)
print(df.head())

• Mitochondrial PLS pipeline:

from BrainNetAnno.mitochondrial_pls_cge import run_mitochondrial_pls_pipeline

best_n, df = run_mitochondrial_pls_pipeline(
    fc_matrix_path="path/to/fc_deviation.csv",      # square matrix CSV, no header
    nt_contrib_csv="path/to/mito_contrib.csv",      # contains 'Region_Pair'
    output_weights_csv="path/to/mito_weights.csv",
)
print(best_n)
print(df.head())

Data Format

• FC deviation/weights CSV: square numeric matrix without header; only upper-triangle non-zero entries are used.
• Gene contribution CSV: must include a Region_Pair column formatted as i-j, where i < j.
• Cell-type markers CSV: columns gene and class (gene symbols are uppercased internally).
• Layer markers Excel: sheet contains columns named t_stat_LayerX (e.g., Layer1..Layer6) and a gene column.

Development & Testing

• Build and install locally:

python -m pip install -U build
python -m build
pip install dist/BrainNetAnno-<version>-py3-none-any.whl

License

MIT