cellink.tl.external.prepare_scprs_data

cellink.tl.external.prepare_scprs_data(adata, labels, gwas_file, prefix, prs_dir, p_thresholds=None, r2_thresholds=None, n_pcs_knn=20, knn_n_neighbors=25, overwrite=False)

Prepare scPRS input data from pre-computed PLINK score files.

Reads PLINK score files generated by the SLURM array job (write_slurm_array_job()) into a (n_samples, n_cells, 21) PRS matrix, builds the kNN graph, and serialises (X, y, edge) as a pickle for run_scprs().

Parameters:

• adata (AnnData) – Reference scATAC-seq AnnData. LSI embedding should be in adata.obsm["X_lsi"] (preferred) or adata.obsm["X_pca"].

• labels (pd.Series) – Binary case/control labels (1=case, 0=control) indexed by donor ID, matching the sample IDs in the PLINK .fam file.

• gwas_file (str or Path) – Post-QC GWAS summary statistics (tab-separated). Only used to determine p-value thresholds if not supplied explicitly.

• prefix (str) – Output prefix for all files.

• prs_dir (str or Path) – Directory containing pre-computed PLINK score files in the layout {prs_dir}/{r2}/{cell}.{p}.profile produced by the SLURM array job. Run write_slurm_array_job() first, then pass the same prs_dir here.

• p_thresholds (list of float, optional) – C+T p-value thresholds. Default: [1e-5, 1e-4, 1e-3, 0.01, 0.05, 0.1, 0.5]

• r2_thresholds (list of float, optional) – C+T LD r² thresholds. Default: [0.1, 0.3, 0.5]

• n_pcs_knn (int, default=20) – Number of LSI/PCA dimensions used for kNN graph.

• knn_n_neighbors (int, default=25) – k for kNN graph (paper default: 25).

• overwrite (bool, default=False) – Re-run even if the output pickle already exists.

Return type:

str

Returns:

str Path to the generated pickle file {prefix}_scprs_data.pkl.