Data Sources

Overview

This document summarizes the spatial transcriptomics, Single-cell RNA-seq, bulk RNA-seq, and RNA-based gene signature data sources used in this study. Datasets include private clinical trial cohorts and publicly available reference and validation datasets, curated to ensure reproducibility, transparency, and consistency with the accompanying spatial transcriptomics manuscript.

Data Sources

Spatial Transcriptomics Data Sources (Discovery Cohort)

Spatial transcriptomic profiling was performed on tumor biopsies from patients enrolled in investigator-initiated immune checkpoint blockade (ICB) trials:

Data type - 10x Genomics Visium spatial transcriptomics from FFPE tissue sections (10 μm)
- Cell Ranger–generated outputs including gene expression matrices, spatial coordinates, and histology images

Access - Restricted due to patient privacy and institutional data-use agreements
- Data are not publicly distributed

Single-Cell RNA-seq Reference Data

Public single-cell RNA-seq datasets were used exclusively as reference atlases for cell-type deconvolution and spatial annotation.

  • Source: Gene Expression Omnibus (GEO) - GSE181919, GSE182227, GSE188737
  • Data type: Whole-transcriptome single-cell RNA-seq
  • Access: Public

Bulk RNA-seq Data (External Validation Cohorts)

Public Validation Datasets

Bulk RNA-seq datasets were used to evaluate the generalizability of spatially informed gene expression signatures.

  • Name: Immune Checkpoint Blockade - RNA-Seq, and Clinical data
  • URL: https://www.orcestra.ca/clinical_icb
  • Access Method: Direct download or programmatic retrieval via API (if applicable)
  • Data Format: MultiAssayExperiment and SummarizedExperiment in R (Bioconductor)
  • Access: Public

Only cohorts meeting predefined inclusion criteria were used.

Private Validation Datasets

In addition to publicly available cohorts, selected private bulk RNA-seq datasets were used for external validation of spatially informed gene expression signatures.

  • Data type: Pre-treatment whole-transcriptome bulk RNA-seq with associated clinical outcome annotations
  • Source: Institutional and collaborative immunotherapy cohorts not publicly released
  • Access: Restricted under institutional review board (IRB) approvals and data-sharing agreements

These datasets were processed and analyzed using the same harmonization, normalization, and scoring procedures applied to public validation cohorts. Due to patient privacy and consent restrictions, raw data are not redistributed in this repository.

Representative private validation cohorts include:

Study / Lead Author Trial / Program Cancer Type Data Repository Access Status
Ravi et al. SU2C-MAR Non-small cell lung cancer dbGaP / EGA Restricted
McDermott et al. IMmotion150 Renal cell carcinoma EGA Restricted
Priestley et al. Hartwig Medical Foundation Pan-cancer HMF Restricted
Rittmeyer et al. OAK (NCT02008227) Non-small cell lung cancer EGA Restricted
Fehrenbacher et al. POPLAR (NCT01903993) Non-small cell lung cancer EGA Restricted

Access to these datasets requires approval from the corresponding data custodians and compliance with institutional and ethical data-use agreements. Cohort-level inclusion, preprocessing, and harmonization procedures are described in the Materials and Methods and Supplementary Data of the manuscript.

Gene Signature Resources

RNA-based gene expression signatures were used for spatial feature scoring, enrichment analysis, and benchmarking.

  • Name: SignatureSets: An R Package for RNA-Based Immuno-Oncology Signatures
  • Version: v1.0
  • URL - IO Signatures: bhklab/SignatureSets
  • Access Method: Direct download or programmatic retrieval via API (if applicable)
  • Data Format: rda, CSV (signatures, metadata)

Signatures were applied using rank-based, enrichment-based, or weighted scoring approaches, depending on the analysis.

Key Clinical Variables

Clinical annotations were harmonized across datasets when applicable. Key variables include:

Variable Description Format Example
patientid Unique patient identifier string GSE12345_P01
age Age at diagnosis integer 63
sex Biological sex factor M/F
cancer_type Primary cancer type string Melanoma
histo Histological classification string Melanoma
treatment_type IO therapy category string PD-1/PD-L1
stage Tumor stage at diagnosis string Stage II
recist RECIST clinical response factor CR/PR/SD/PD
response Clinical benefit status (e.g., response) string R/NR
survival_time_os Overall survival time (months) numeric 21.3
event_occurred_os Overall survival event (1 = death) binary 1
survival_time_pfs Progression-free survival time (months) numeric 18.2
event_occurred_pfs Progression-free survival event (1 = progression) binary 1
survival_unit Unit of survival time string months