Steve Rozen
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Steve Rozen has a total of 1000 co-authors. Here are the top co-authors by number of shared publications:
- Bin Tean Teh
Cancer and Stem Cell Biology Programme, Duke-NUS Medical School, Singapore 169857, Singapore
78 shared publications - Patrick Tan
Cancer and Stem Cell Biology Program, Duke–NUS Graduate Medical School, Singapore, Singapore
62 shared publications - Choon Kiat Ong
Duke-NUS Medical School, Singapore, Singapore
61 shared publications Anna Gan
52 shared publications- Willie Yu
Centre for Biomedical Data Science, Duke-NUS Medical School, Singapore 169857, Singapore
44 shared publications - Ioana Cutcutache
Neuroscience and Behavioral Disorders, Duke-NUS Graduate Medical School, 8 College Road, Singapore, 169857
42 shared publications
Latest Publications
- 10 November 2026
- Published by Cold Spring Harbor Laboratory in Genome Research
Derivation of doxorubicin mutational signatures in cancer
Doxorubicin is a chemotherapeutic agent used for the treatment of solid and hematopoietic malignancies. Doxorubicin acts as a DNA intercalator and inhibits the topoisomerase II enzyme, leading to double-stranded breaks. We hypothesize that doxorubicin mutagenesis leads to a nonrandom pattern of somatic mutations on the genome, termed a mutational signature. We expose cells to doxorubicin using two different methodologies: (1) continuous exposure of a constant dose of doxorubicin and whole-genome sequencing of single-cell clones (previously published methodology) and (2) pulsed exposure of escalating doses of doxorubicin and whole-genome sequencing of a bulk population. We first confirm that the methodologies do not yield differing results through previously published cisplatin-exposed single-cell clone genome data compared with the dose escalation and bulk genome data. We then derive two doxorubicin signatures. The first, SBS-doxorubicin-1, is defined by an abundance of C > A mutations. The second, SBS-doxorubicin-2, is defined by a higher occurrence of T > C mutations. We then assess for the presence of our signatures within the genomes of patients treated with a doxorubicin-containing regimen. Across 10 bulk immune sorted samples from three patients with follicular lymphoma/diffuse large B cell lymphoma, as well as two metastatic samples from one patient with renal medullary carcinoma, we are able to detect both signatures in 83% of samples, with remaining samples harboring either signature. In summary, we experimentally derive two doxorubicin mutational signatures and demonstrate their presence in patient samples.
- 10 November 2026
- Published by Cold Spring Harbor Laboratory in Genome Research
Derivation of doxorubicin mutational signatures in cancer
Doxorubicin is a chemotherapeutic agent used for the treatment of solid and hematopoietic malignancies. Doxorubicin acts as a DNA intercalator and inhibits the topoisomerase II enzyme, leading to double-stranded breaks. We hypothesize that doxorubicin mutagenesis leads to a nonrandom pattern of somatic mutations on the genome, termed a mutational signature. We expose cells to doxorubicin using two different methodologies: (1) continuous exposure of a constant dose of doxorubicin and whole-genome sequencing of single-cell clones (previously published methodology) and (2) pulsed exposure of escalating doses of doxorubicin and whole-genome sequencing of a bulk population. We first confirm that the methodologies do not yield differing results through previously published cisplatin-exposed single-cell clone genome data compared with the dose escalation and bulk genome data. We then derive two doxorubicin signatures. The first, SBS-doxorubicin-1, is defined by an abundance of C > A mutations. The second, SBS-doxorubicin-2, is defined by a higher occurrence of T > C mutations. We then assess for the presence of our signatures within the genomes of patients treated with a doxorubicin-containing regimen. Across 10 bulk immune sorted samples from three patients with follicular lymphoma/diffuse large B cell lymphoma, as well as two metastatic samples from one patient with renal medullary carcinoma, we are able to detect both signatures in 83% of samples, with remaining samples harboring either signature. In summary, we experimentally derive two doxorubicin mutational signatures and demonstrate their presence in patient samples.
- Open Access
- 28 May 2025
- Published by Elsevier in Journal of the National Cancer Center
- Vol. 5 (5) , 515-523
Transcriptomic predictors of prostate cancer recurrence following focal cryotherapy: a pooled analysis of phase II trial and prospective cohort data
- 28 May 2025
- Published by Elsevier in Journal of the National Cancer Center
- Vol. 5 (5) , 515-523
Open AccessTranscriptomic predictors of prostate cancer recurrence following focal cryotherapy: a pooled analysis of phase II trial and prospective cohort data
- Open Access
- 22 November 2024
- Published by Oxford University Press (OUP) in Briefings in Bioinformatics
- Vol. 26 (1)
Benchmarking 13 tools for mutational signature attribution, including a new and improved algorithm
Mutational signatures are characteristic patterns of mutations caused by endogenous mutational processes or by exogenous mutational exposures. There has been little benchmarking of approaches for determining which signatures are present in a sample and estimating the number of mutations due to each signature. This problem is referred to as “signature attribution.” We show that there are often many combinations of signatures that can reconstruct the patterns of mutations in a sample reasonably well, even after encouraging sparse solutions. We benchmarked 13 approaches to signature attribution, including a new approach called Presence Attribute Signature Activity (PASA), on large synthetic data sets (2700 synthetic samples in total). These data sets recapitulated the single-base, insertion–deletion, and doublet-base mutational signature repertoires of nine cancer types. For single-base substitution mutations, PASA and MuSiCal outperformed other approaches on all the cancer types combined. However, the ranking of approaches varied by cancer type. For doublet-base substitutions and small insertions and deletions, while PASA outperformed the other approaches in most of the nine cancer types, the ranking of approaches again varied by cancer type. We believe that this variation reflects inherent difficulties in signature attribution. These difficulties stem from the fact that there are often many attributions that can reasonably explain the pattern of mutations in a sample and from the combinatorial search space due to the need to impose sparsity. Tables herein can provide guidance on the selection of mutational signature attribution approaches that are best suited to particular cancer types and study objectives.
- 22 November 2024
- Published by Oxford University Press (OUP) in Briefings in Bioinformatics
- Vol. 26 (1)
Open AccessBenchmarking 13 tools for mutational signature attribution, including a new and improved algorithm
Mutational signatures are characteristic patterns of mutations caused by endogenous mutational processes or by exogenous mutational exposures. There has been little benchmarking of approaches for determining which signatures are present in a sample and estimating the number of mutations due to each signature. This problem is referred to as “signature attribution.” We show that there are often many combinations of signatures that can reconstruct the patterns of mutations in a sample reasonably well, even after encouraging sparse solutions. We benchmarked 13 approaches to signature attribution, including a new approach called Presence Attribute Signature Activity (PASA), on large synthetic data sets (2700 synthetic samples in total). These data sets recapitulated the single-base, insertion–deletion, and doublet-base mutational signature repertoires of nine cancer types. For single-base substitution mutations, PASA and MuSiCal outperformed other approaches on all the cancer types combined. However, the ranking of approaches varied by cancer type. For doublet-base substitutions and small insertions and deletions, while PASA outperformed the other approaches in most of the nine cancer types, the ranking of approaches again varied by cancer type. We believe that this variation reflects inherent difficulties in signature attribution. These difficulties stem from the fact that there are often many attributions that can reasonably explain the pattern of mutations in a sample and from the combinatorial search space due to the need to impose sparsity. Tables herein can provide guidance on the selection of mutational signature attribution approaches that are best suited to particular cancer types and study objectives.
- Open Access
- 14 June 2024
- Published by American Association for Cancer Research (AACR)
Supplementary Tables from Oncogene-Driven Non–Small Cell Lung Cancers in Patients with a History of Smoking Lack Smoking-Induced Mutations
Supplementary Tables from Oncogene-Driven Non–Small Cell Lung Cancers in Patients with a History of Smoking Lack Smoking-Induced Mutations
- 14 June 2024
- Published by American Association for Cancer Research (AACR)
Open AccessSupplementary Tables from Oncogene-Driven Non–Small Cell Lung Cancers in Patients with a History of Smoking Lack Smoking-Induced Mutations
Supplementary Tables from Oncogene-Driven Non–Small Cell Lung Cancers in Patients with a History of Smoking Lack Smoking-Induced Mutations
- Open Access
- 14 June 2024
- Published by American Association for Cancer Research (AACR)
Supplementary Tables from Oncogene-Driven Non–Small Cell Lung Cancers in Patients with a History of Smoking Lack Smoking-Induced Mutations
Supplementary Tables from Oncogene-Driven Non–Small Cell Lung Cancers in Patients with a History of Smoking Lack Smoking-Induced Mutations
- 14 June 2024
- Published by American Association for Cancer Research (AACR)
Open AccessSupplementary Tables from Oncogene-Driven Non–Small Cell Lung Cancers in Patients with a History of Smoking Lack Smoking-Induced Mutations
Supplementary Tables from Oncogene-Driven Non–Small Cell Lung Cancers in Patients with a History of Smoking Lack Smoking-Induced Mutations
- Open Access
Benchmarking 13 tools for mutational signature attribution, including a new and improved algorithm
Mutational signatures are characteristic patterns of mutations caused by endogenous mutational processes or by exogenous mutational exposures. There has been little benchmarking of approaches for determining which signatures are present in a sample and estimating the number of mutations due to each signature. This problem is referred to as “signature attribution”. We show that there are often many combinations of signatures that can reconstruct the patterns of mutations in a sample reasonably well, even after encouraging sparse solutions. We benchmarked thirteen approaches to signature attribution, including a new approach called Presence Attribute Signature Activity (PASA), on large synthetic data sets (2,700 synthetic samples in total). These data sets recapitulated the single-base, insertion-deletion, and doublet-base mutational signature repertoires of 9 cancer types. For single-base substitution mutations, PASA and MuSiCal outperformed other approaches on all the cancer types combined. However, the ranking of approaches varied by cancer type. For doublet-base substitutions and small insertions and deletions, while PASA outperformed the other approaches in most of the nine cancer types, the ranking of approaches again varied by cancer type. We believe this variation reflects inherent difficulties in signature attribution. These difficulties stem from the fact that there are often many attributions that can reasonably explain the pattern of mutations in a sample and from the combinatorial search space due to the need to impose sparsity. Tables herein can provide guidance on the selection of mutational signature attribution approaches that are best suited to particular cancer types and study objectives.
Open AccessBenchmarking 13 tools for mutational signature attribution, including a new and improved algorithm
Mutational signatures are characteristic patterns of mutations caused by endogenous mutational processes or by exogenous mutational exposures. There has been little benchmarking of approaches for determining which signatures are present in a sample and estimating the number of mutations due to each signature. This problem is referred to as “signature attribution”. We show that there are often many combinations of signatures that can reconstruct the patterns of mutations in a sample reasonably well, even after encouraging sparse solutions. We benchmarked thirteen approaches to signature attribution, including a new approach called Presence Attribute Signature Activity (PASA), on large synthetic data sets (2,700 synthetic samples in total). These data sets recapitulated the single-base, insertion-deletion, and doublet-base mutational signature repertoires of 9 cancer types. For single-base substitution mutations, PASA and MuSiCal outperformed other approaches on all the cancer types combined. However, the ranking of approaches varied by cancer type. For doublet-base substitutions and small insertions and deletions, while PASA outperformed the other approaches in most of the nine cancer types, the ranking of approaches again varied by cancer type. We believe this variation reflects inherent difficulties in signature attribution. These difficulties stem from the fact that there are often many attributions that can reasonably explain the pattern of mutations in a sample and from the combinatorial search space due to the need to impose sparsity. Tables herein can provide guidance on the selection of mutational signature attribution approaches that are best suited to particular cancer types and study objectives.
- Open Access
- 8 April 2024
- Published by American Association for Cancer Research (AACR) in Cancer Research
- Vol. 84 (12) , 2009-2020
Oncogene-Driven Non–Small Cell Lung Cancers in Patients with a History of Smoking Lack Smoking-Induced Mutations
Non–small cell lung cancers (NSCLC) in nonsmokers are mostly driven by mutations in the oncogenes EGFR, ERBB2, and MET and fusions involving ALK and RET. In addition to occurring in nonsmokers, alterations in these “nonsmoking-related oncogenes” (NSRO) also occur in smokers. To better understand the clonal architecture and genomic landscape of NSRO-driven tumors in smokers compared with typical-smoking NSCLCs, we investigated genomic and transcriptomic alterations in 173 tumor sectors from 48 NSCLC patients. NSRO-driven NSCLCs in smokers and nonsmokers had similar genomic landscapes. Surprisingly, even in patients with prominent smoking histories, the mutational signature caused by tobacco smoking was essentially absent in NSRO-driven NSCLCs, which was confirmed in two large NSCLC data sets from other geographic regions. However, NSRO-driven NSCLCs in smokers had higher transcriptomic activities related to the regulation of the cell cycle. These findings suggest that, whereas the genomic landscape is similar between NSRO-driven NSCLC in smokers and nonsmokers, smoking still affects the tumor phenotype independently of genomic alterations. Significance: Non-small cell lung cancers driven by nonsmoking-related oncogenes do not harbor genomic scars caused by smoking regardless of smoking history, indicating that the impact of smoking on these tumors is mainly nongenomic.
- 8 April 2024
- Published by American Association for Cancer Research (AACR) in Cancer Research
- Vol. 84 (12) , 2009-2020
Open AccessOncogene-Driven Non–Small Cell Lung Cancers in Patients with a History of Smoking Lack Smoking-Induced Mutations
Non–small cell lung cancers (NSCLC) in nonsmokers are mostly driven by mutations in the oncogenes EGFR, ERBB2, and MET and fusions involving ALK and RET. In addition to occurring in nonsmokers, alterations in these “nonsmoking-related oncogenes” (NSRO) also occur in smokers. To better understand the clonal architecture and genomic landscape of NSRO-driven tumors in smokers compared with typical-smoking NSCLCs, we investigated genomic and transcriptomic alterations in 173 tumor sectors from 48 NSCLC patients. NSRO-driven NSCLCs in smokers and nonsmokers had similar genomic landscapes. Surprisingly, even in patients with prominent smoking histories, the mutational signature caused by tobacco smoking was essentially absent in NSRO-driven NSCLCs, which was confirmed in two large NSCLC data sets from other geographic regions. However, NSRO-driven NSCLCs in smokers had higher transcriptomic activities related to the regulation of the cell cycle. These findings suggest that, whereas the genomic landscape is similar between NSRO-driven NSCLC in smokers and nonsmokers, smoking still affects the tumor phenotype independently of genomic alterations. Significance: Non-small cell lung cancers driven by nonsmoking-related oncogenes do not harbor genomic scars caused by smoking regardless of smoking history, indicating that the impact of smoking on these tumors is mainly nongenomic.
- Top Cited Papers
- 14 February 2024
- Published by Springer Nature in Nature
- Vol. 627 (8004) , 586-593
Deep whole-genome analysis of 494 hepatocellular carcinomas
Over half of hepatocellular carcinoma (HCC) cases diagnosed worldwide are in China1,2,3. However, whole-genome analysis of hepatitis B virus (HBV)-associated HCC in Chinese individuals is limited4,5,6,7,8, with current analyses of HCC mainly from non-HBV-enriched populations9,10. Here we initiated the Chinese Liver Cancer Atlas (CLCA) project and performed deep whole-genome sequencing (average depth, 120×) of 494 HCC tumours. We identified 6 coding and 28 non-coding previously undescribed driver candidates. Five previously undescribed mutational signatures were found, including aristolochic-acid-associated indel and doublet base signatures, and a single-base-substitution signature that we termed SBS_H8. Pentanucleotide context analysis and experimental validation confirmed that SBS_H8 was distinct to the aristolochic-acid-associated SBS22. Notably, HBV integrations could take the form of extrachromosomal circular DNA, resulting in elevated copy numbers and gene expression. Our high-depth data also enabled us to characterize subclonal clustered alterations, including chromothripsis, chromoplexy and kataegis, suggesting that these catastrophic events could also occur in late stages of hepatocarcinogenesis. Pathway analysis of all classes of alterations further linked non-coding mutations to dysregulation of liver metabolism. Finally, we performed in vitro and in vivo assays to show that fibrinogen alpha chain (FGA), determined as both a candidate coding and non-coding driver, regulates HCC progression and metastasis. Our CLCA study depicts a detailed genomic landscape and evolutionary history of HCC in Chinese individuals, providing important clinical implications.
- 14 February 2024
- Published by Springer Nature in Nature
- Vol. 627 (8004) , 586-593
Top Cited PapersDeep whole-genome analysis of 494 hepatocellular carcinomas
Over half of hepatocellular carcinoma (HCC) cases diagnosed worldwide are in China1,2,3. However, whole-genome analysis of hepatitis B virus (HBV)-associated HCC in Chinese individuals is limited4,5,6,7,8, with current analyses of HCC mainly from non-HBV-enriched populations9,10. Here we initiated the Chinese Liver Cancer Atlas (CLCA) project and performed deep whole-genome sequencing (average depth, 120×) of 494 HCC tumours. We identified 6 coding and 28 non-coding previously undescribed driver candidates. Five previously undescribed mutational signatures were found, including aristolochic-acid-associated indel and doublet base signatures, and a single-base-substitution signature that we termed SBS_H8. Pentanucleotide context analysis and experimental validation confirmed that SBS_H8 was distinct to the aristolochic-acid-associated SBS22. Notably, HBV integrations could take the form of extrachromosomal circular DNA, resulting in elevated copy numbers and gene expression. Our high-depth data also enabled us to characterize subclonal clustered alterations, including chromothripsis, chromoplexy and kataegis, suggesting that these catastrophic events could also occur in late stages of hepatocarcinogenesis. Pathway analysis of all classes of alterations further linked non-coding mutations to dysregulation of liver metabolism. Finally, we performed in vitro and in vivo assays to show that fibrinogen alpha chain (FGA), determined as both a candidate coding and non-coding driver, regulates HCC progression and metastasis. Our CLCA study depicts a detailed genomic landscape and evolutionary history of HCC in Chinese individuals, providing important clinical implications.