Data and Codes for Pan-cancer classification of single cells in the tumour microenvironment
Bibliographic record
Abstract
Data for Pan-cancer classification of single cells in the tumour microenvironment. Most files are saved in .RDS format and can be loaded in R using the readRDS function training_files_for_scATOMIC_core.tar.gz contains the count matrices, metadata files, and scripts used to train the core scATOMIC model. The reference_datasets folder contains training matrices and metadata that were used to derive marker genes and balanced training matrices using the get_markers_per_cell_type_per_layer.R and generate_balanced_training.R scripts. The markers folder contains the results of differential gene expression between cell types at each layer. The balanced_training folder contains the class balanced matrices used to train each layer in the train_classifiers.R script. The classifier_outputs folder contains the output models and gene lists used. external_validation_files.tar.gz contains the count matrices and results used in the external validation of scATOMIC. harmonized_results_matrix_and_metadata_all_patients_all_cells.RDS is a matrix of all cells with their associated metadata, scATOMIC annotations and annotations with other tools. gold_standard_cells_matrix_for_fig2bc.RDS is a filtered versions of cells with a more confident ground truth that were used for external validation in figure_2bc_script.R. We also provide files split by each patient: split_by_patient_matrices contains each count matrix for each patient tested. prediction_list_files contains the results of run_scATOMIC() for each patient. scATOMIC_annotations_per_sample contains the summarized results matrix for each individual patient. Figure3.tar.gz contains the script used to compare scATOMIC's cancer signature scoring mode to copykat CNV inference. The script requires files from external_validation_files.tar.gz. Pal_et_al_breast_cancer_data.tar.gz contains the results of scATOMIC in each breast cancer subtype and the script to analyze the different subtype annotations as a pie chart. metastatic_figure.tar.gz contains the results of scATOMIC prediction of metastatic disease origin as well as the script to run the analysis. scATOMIC_1.1.0.tar.gz contains the R package (version used in manuscript).
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.016 |
| Meta-epidemiology (narrow) | 0.002 | 0.001 |
| Meta-epidemiology (broad) | 0.002 | 0.002 |
| Bibliometrics | 0.005 | 0.007 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.003 | 0.002 |
| Open science | 0.003 | 0.002 |
| Research integrity | 0.002 | 0.003 |
| Insufficient payload (model declined to judge) | 0.314 | 0.188 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".