Deep Learning-Based Inference of Mutational Signature Exposures from Histology Across Tumour Types
4 affiliations
- 1CMC Institute for Basic Medical Science, The Catholic Medical Center, The Catholic University of Korea, Seoul 06591, Republic of Korea
- 2Department of Medical Informatics, College of Medicine, The Catholic University of Korea, Seoul 06591, Republic of Korea
- 3Graduate School, The Catholic University of Korea, Seoul 06591, Republic of Korea
- 4Cancer Research Institute, College of Medicine, The Catholic University of Korea, Seoul 06591, Republic of Korea
Abstract
Background/Objectives: Mutational signatures reflect the processes that shape cancer genomes. We investigated whether continuous mutational-signature exposures could be inferred from routine histology across tumour types. Methods: We developed whole-slide image regression models for 30 single-base substitution (SBS) signatures in 2741 The Cancer Genome Atlas (TCGA) patients across seven tumour types and evaluated them in 580 Clinical Proteomic Tumor Analysis Consortium (CPTAC) patients. RetCCL mean-feature ridge regression and RetCCL/CTransPath attention-based multiple-instance learning were combined by fixed unweighted averaging. Final ensemble outputs were standardised SBS scores. Patient-disjoint evaluation, full-panel false discovery rate (FDR) adjustment and tumour-labelled slide sensitivity were performed. Results: Among 185 pairs with correlations available in both cohorts, 73 showed positive correlations in both cohorts. Fourteen had nominal positive-direction p < 0.05 in both, including breast cancer–SBS3 (TCGA r = 0.289, CPTAC r = 0.215) and lung adenocarcinoma–SBS4 (r = 0.128 and 0.263). Cohort-specific FDR-adjusted values are provided for all 210 pairs. All 14 associations remained positive after restricting inputs to tumour-labelled slides. The SBS3 score correlated with continuous homologous recombination deficiency (r = 0.332). Exploratory TCGA SBS3, SBS6 and SBS10 scores yielded areas under the receiver operating characteristic curve of 0.693, 0.619 and 0.628 for homologous recombination deficiency-high, microsatellite instability and qualifying POLE/POLD1 mutations. External MSI and polymerase AUCs were 0.573 and 0.680, respectively. Conclusions: Histology-derived scores were associated with selected genomic SBS exposures in both cohorts. Further studies are needed to identify the tissue features underlying these associations and assess their clinical relevance.
Related
- Mutational signature landscape of pancreatic cancer genomes from nine countries
- Human cancer genomes harbor the mutational signature of tobacco-specific nitrosamines NNN and NNK
- A healthy donor kidney aging signature estimable from PAS histology and its association with progression of diabetic kidney disease
- Machine learning-guided identification of determinants of PD-L1 expression and tumor mutational burden in lung squamous cell carcinoma.
- RNA-guided contrastive learning enhances patient-level prediction from histology
BibTeX
@article{sim2026deep,
title = {{Deep Learning-Based Inference of Mutational Signature Exposures from Histology Across Tumour Types}},
author = {Byeong Chang Sim and JunSik Kim and Jiyeon Park and YunJi Kim and JungHyun Park and Taehoon Ko and Ji-Won Chun and Tae-Min Kim},
journal = {Cancers},
year = {2026},
month = oct,
doi = {10.3390/cancers18213446},
url = {https://www.mdpi.com/2072-6694/18/21/3446},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - Deep Learning-Based Inference of Mutational Signature Exposures from Histology Across Tumour Types
AU - Byeong Chang Sim
AU - JunSik Kim
AU - Jiyeon Park
AU - YunJi Kim
AU - JungHyun Park
AU - Taehoon Ko
AU - Ji-Won Chun
AU - Tae-Min Kim
PY - 2026
DA - 2026/10/25
JO - Cancers
DO - 10.3390/cancers18213446
UR - https://www.mdpi.com/2072-6694/18/21/3446
AB - Background/Objectives: Mutational signatures reflect the processes that shape cancer genomes. We investigated whether continuous mutational-signature exposures could be inferred from routine histology across tumour types. Methods: We developed whole-slide image regression models for 30 single-base substitution (SBS) signatures in 2741 The Cancer Genome Atlas (TCGA) patients across seven tumour types and evaluated them in 580 Clinical Proteomic Tumor Analysis Consortium (CPTAC) patients. RetCCL mean-feature ridge regression and RetCCL/CTransPath attention-based multiple-instance learning were combined by fixed unweighted averaging. Final ensemble outputs were standardised SBS scores. Patient-disjoint evaluation, full-panel false discovery rate (FDR) adjustment and tumour-labelled slide sensitivity were performed. Results: Among 185 pairs with correlations available in both cohorts, 73 showed positive correlations in both cohorts. Fourteen had nominal positive-direction p < 0.05 in both, including breast cancer–SBS3 (TCGA r = 0.289, CPTAC r = 0.215) and lung adenocarcinoma–SBS4 (r = 0.128 and 0.263). Cohort-specific FDR-adjusted values are provided for all 210 pairs. All 14 associations remained positive after restricting inputs to tumour-labelled slides. The SBS3 score correlated with continuous homologous recombination deficiency (r = 0.332). Exploratory TCGA SBS3, SBS6 and SBS10 scores yielded areas under the receiver operating characteristic curve of 0.693, 0.619 and 0.628 for homologous recombination deficiency-high, microsatellite instability and qualifying POLE/POLD1 mutations. External MSI and polymerase AUCs were 0.573 and 0.680, respectively. Conclusions: Histology-derived scores were associated with selected genomic SBS exposures in both cohorts. Further studies are needed to identify the tissue features underlying these associations and assess their clinical relevance.
SN - 2072-6694
ER -