Related Experiment Video
Updated: Aug 5, 2026
Discovery of Driver Genes in Colorectal HT29-derived Cancer Stem-Like Tumorspheres
Published on: July 22, 2020
scOPE identifies which driver-associated expression programs transfer from bulk tumors to single cells
Abstract:
Single-cell RNA sequencing (scRNA-seq) resolves the phenotypic heterogeneity of tumors but rarely observes the somatic mutations that drive it: a variant is legible only where its gene is expressed, the mutant allele is transcribed, and reads span the variant site, so an absent variant read is fundamentally ambiguous. Bulk tumor cohorts have the opposite profile-matched genotype and expression for hundreds of patients, but no cellular resolution. We present scOPE ( s ingle- c ell O ncological P rediction E xplorer), which learns cancer-specific, driver-associated expression axes from bulk tumors, freezes them, and projects single-cell transcriptomes onto the fixed axes without refitting to the target cohort. Our central finding is that this transfer is selective rather than general: of 158 audited driver-cancer models across seven malignancies, 102 met predefined claim-safety criteria and only 11 reached out-of-fold AUROC ≥ 0.90, led by acute myeloid leukemia (AML) NPM1 (0.971), glioblastoma IDH1 (0.963), and pancreatic adenocarcinoma KRAS (0.944). Determining which programs transfer therefore becomes the central task. We address it with a ground-truth-free confidence score-integrating bulk transferability, spatial coherence, score concentration, and copy-number (CNV) agreement-that within AML ranked the three independently supported programs above the remainder (AUROC 0.85 across 12 truth-evaluable drivers, of which three were supported), a triage signal rather than a validated genotype classifier. Against expressed-mutation labels, cell-state-residual scores were enriched in mutant-labeled cells for NPM1 , TP53 , and DNMT3A , and the NPM1 separation survived aggregation to patients. Critically, matched genotype-score maps show that even supported programs occupy restricted transcriptional subspaces rather than uniformly marking mutation-positive tumors, and the NPM1 program contracted during treatment across multiple patients. Transferred scores tracked inferred CNV burden yet also resolved discordant malignant populations invisible to aneuploidy alone. scOPE does not call alleles; it recovers continuous, mutation-associated transcriptional axes from existing scRNA-seq data, together with explicit diagnostics for when that reading should be withheld.
Insights
Single-cell RNA sequencing (scRNA-seq) can now infer cancer driver mutations by learning expression patterns from bulk tumors. The scOPE tool selectively transfers these patterns to single-cell data, identifying specific mutations like NPM1 in AML.
Area of Science:
- Oncology
- Genomics
- Bioinformatics
Background:
- Single-cell RNA sequencing (scRNA-seq) offers high resolution of tumor heterogeneity but struggles to detect somatic mutations due to expression and transcription requirements.
- Bulk tumor sequencing provides matched genotype and expression data but lacks cellular resolution.
- Bridging this gap is crucial for understanding mutation-driven tumor evolution.
Purpose of the Study:
- To develop and validate a computational tool, scOPE (single-cell Oncological Prediction Explorer), for inferring cancer driver mutations from scRNA-seq data.
- To assess the transferability and accuracy of cancer-specific expression axes learned from bulk tumors onto single-cell transcriptomes.
- To establish confidence metrics for evaluating the reliability of inferred mutation-associated transcriptional programs.
Main Methods:
- scOPE learns cancer-specific, driver-associated expression axes from bulk tumor data.
- These learned axes are then projected onto single-cell transcriptomes without refitting.
- A confidence score integrating bulk transferability, spatial coherence, score concentration, and copy-number variation (CNV) agreement is used to assess program reliability.
Main Results:
- The scOPE tool demonstrated selective transferability of driver-associated expression programs across seven malignancies, with 102 out of 158 models meeting safety criteria.
- High accuracy (AUROC ≥ 0.90) was achieved for specific driver-cancer pairs, notably AML NPM1 (0.971), glioblastoma IDH1 (0.963), and pancreatic adenocarcinoma KRAS (0.944).
- A confidence score effectively prioritized validated programs in AML and revealed mutation-associated transcriptional programs enriched in mutant-labeled cells (NPM1, TP53, DNMT3A), with NPM1 separation persisting at the patient level.
Conclusions:
- scOPE enables the recovery of continuous, mutation-associated transcriptional axes from existing scRNA-seq data, providing valuable insights into tumor biology.
- The tool offers explicit diagnostics to indicate when inferred transcriptional programs should be interpreted with caution.
- Inferred scores track CNV burden and can resolve discordant malignant populations, offering a new dimension for cancer subtyping and treatment monitoring.
