| # P2M.db Analysis Code |
|
|
| - `analysis/`: complete analysis scripts partitioned into `scRNA-seq_code/` and `Bulk-code/`, documentation and MIT license. |
| - [validation/](validation/): matrix QC, five-cohort annotation comparisons, marker checks and manuscript figures. |
| - [validation/parameters.csv](validation/parameters.csv): shared workflow settings; recorded accession-specific settings remain in each dataset directory. |
| - `read_data.R`: example for reading expression data and specimen metadata. |
| - `parameters.py`: general formatter for processing_parameters.csv parameter group tables. |
| - `export_expression.R`: general export utility for RDS expression matrices and metadata. |
|
|
| ## Read the data |
|
|
| Requires R and SeuratObject. |
|
|
| ```sh |
| Rscript code/read_data.R scRNA-seq/GSE264205/GSE264205.rds |
| Rscript code/read_data.R bulk/GSE205209/expression.rds |
| ``` |
|
|
| For a bulk object containing multiple matrices, supply an exact matrix name as |
| argument 2. If sample IDs recur across platforms, supply the platform as argument 3. |
| The single-cell example uses specimen_role to select tumor tissue. Malignant ascites, other fluid specimens, normal and unresolved roles are excluded even when their coarse sample_class is Metastasis. Bulk selections use the recorded specimen class. |
| TCGA clinical M0/M1 is distinct from specimen class. Annotation methods remain |
| independent. Inspect the cohort metadata and comparison definitions before analysis. |
|
|
| ## Analysis scripts |
|
|
| Analysis pipelines are partitioned into dedicated subdirectories under `analysis/`: |
|
|
| ### Single-cell RNA-seq scripts (`analysis/scRNA-seq_code/`) |
| All single-cell analysis and downstream formatting scripts are organized under `code/analysis/scRNA-seq_code/`: |
| - `single_cell.R`: original single-cell pipeline (QC, normalization, Harmony integration, SingleR annotation, CellChat, Monocle3, clusterProfiler). |
| - `single_cell_analysis.R`: multi-stage comparative analysis pipeline for added single-cell cohorts. |
| ```sh |
| Rscript code/analysis/scRNA-seq_code/single_cell_analysis.R GSE203067 prepare |
| ``` |
| - `annotate_single.R`: automated and reference-based cell annotation using SingleR and scType marker scoring. |
| ```sh |
| Rscript code/analysis/scRNA-seq_code/annotate_single.R GSE203067 |
| ``` |
| - `doublets.py`: DoubletDetection wrapper script for identifying doublet cells. |
| ```sh |
| python code/analysis/scRNA-seq_code/doublets.py |
| ``` |
| - `format_deg.R`: cell-type differential expression table formatting. |
| - `format_enrichment.R`: cell-type functional enrichment table formatting. |
| - `format_gene_singler.R`: SingleR cell-type marker expression matrix formatting. |
| - `format_gene_sctype.R`: scType cell-type marker expression matrix formatting. |
|
|
| ### Bulk transcriptome scripts (`analysis/Bulk-code/`) |
| All bulk transcriptome analysis and preprocessing pipelines are organized under `code/analysis/Bulk-code/`: |
| - `bulk_geo.R`: bulk GEO microarray and RNA-seq analysis pipeline (limma, GEOquery). |
| - `bulk_tcga.R`: bulk TCGA pan-cancer primary-versus-metastasis processing pipeline. |
| - `bulk_counts.R`: author count matrix ingestion, QC, and packaging into standardized RDS objects. |
| ```sh |
| Rscript code/analysis/Bulk-code/bulk_counts.R INPUT_DIR OUTPUT_DIR GSE317536 package |
| ``` |
|
|
| ### General utilities (top-level `code/`) |
| Top-level general dataset utilities, shared databases, and benchmark code remain at the `code/` root: |
| - `read_data.R`: example for reading expression matrices and specimen metadata. |
| - `parameters.py`: parameter table group formatter. |
| - `export_expression.R`: expression matrix export utility. |
| ```sh |
| Rscript code/export_expression.R scRNA-seq/GSE264205/GSE264205.rds OUTPUT_DIR |
| ``` |
| - `resources/`: reference gene sets and marker databases. |
| - `validation/`: benchmark validation, quality control, and paper figure code. |
|
|
| Source: https://github.com/bigdata-lrkai/P2M.Rcode |
| Commit: `11530593102b06d74ad7199a940adcbb234015f1`. |
| Script filenames have been standardized; their contents and license are preserved. |
| The original scripts retain historical paths, manual grouping and known formatting |
| limitations. Code availability does not certify that all historical analyses were |
| rerun. All files can be downloaded independently from Hugging Face. |
|
|
| ## Parameter table format |
|
|
| Each accession's processing_parameters.csv has two columns, function and parameters, |
| with one function per row. The formatter preserves individual values and can read |
| either the original three-column layout or the grouped layout. |
| |
| ```sh |
| python code/parameters.py --group INPUT.csv OUTPUT.csv |
| ``` |
| |
| Requires Python 3 and the standard library. The output must not already exist. |
| Functions marked `(workflow reference)` describe the original workflow rather than |
| recovered accession-specific execution logs. Recorded values take precedence. |
| |
| ## TCGA survival analyses |
| |
| Overall-survival inputs, all tested gene models and model diagnostics for 18 TCGA |
| cohorts are provided in each `bulk/TCGA_*/survival/` directory; the combined gene |
| results are in `analysis/survival.csv.gz`. Death and last-follow-up times are |
| retained, with one tumor profile per patient. Analyses use clinical M0/M1 groups, |
| which are distinct from specimen tissue origin. Nominal and BH-adjusted Cox |
| P values and proportional-hazards diagnostics are reported separately. Website |
| Kaplan-Meier curves use a median expression split and a log-rank test. |
| |
| The current survival workflow is in |
| [`validation/survival/`](validation/survival/). GEO survival files retain |
| their source-specific endpoints and historical subsets; the TCGA update does not |
| assert that those GEO analyses were reproduced. |
| |
