Abstract
Purpose
Skin cutaneous melanoma (SKCM) is a highly aggressive melanocytic carcinoma whose high heterogeneity and complex etiology make its prognosis difficult to predict. This study aimed to construct a risk subtype typing model for SKCM.
Methods
The study proposes a deep learning framework combining early fusion feature autoencoder (AE) and late fusion feature AE for risk subtype prediction of SKCM. The deep learning framework integrates mRNA, miRNA, and DNA methylation data of SKCM patients from The Cancer Genome Atlas (TCGA), and clusters the screened multi-omics features associated with survival prognosis to identify risk subtypes. Differential expression analysis and functional enrichment analysis were performed between risk subtypes, while SVM classifiers were constructed between differentially expressed genes (DEGs) obtained by Least Absolute Shrinkage and Selection Operator (LASSO) logistic regression screening and risk subtype labels inferred from multi-omics data, and the predictive robustness of risk subtypes inferred from the risk subtype classification prediction model was validated using two independent datasets.
Results
The deep learning framework that combined early fusion feature AE with late fusion feature AE distinguished the two best risk subtypes compared to the multi-omics integration approach with single strategy AE or PCA. A promising C-index (C-index = 0.748) and a significant difference in survival (log-rank P value = 4.61 × 10–9) were found between the identified risk subtypes. The DEGs with the top significance values together with differentially expressed miRNAs provided the biological interpretation of risk subtypes on SKCM. Finally, the framework was applied to predict risk subtypes in two independent test datasets of SKCM patients, all of which showed good predictive power (C-index > 0.680) and significant survival differences (log-rank P value < 0.01).
Conclusion
The SKCM risk subtypes identified by integrating multi-omics data based on deep learning can not only improve the understanding of the molecular mechanisms of SKCM, but also provide clinicians with assistance in treatment decisions.
Similar content being viewed by others
Data availability
All data can be retrieved from the TCGA database (https://portal.gdc.cancer.gov) and the GEO database (https://www.ncbi.nlm.nih.gov).
References
Avogadri F, Zappasodi R, Yang A et al (2014) Combination of Alphavirus replicon particle-based vaccination with immunomodulatory antibodies: therapeutic activity in the B16 melanoma mouse model and immune correlates. Cancer Immunol Res 2:448–458. https://doi.org/10.1158/2326-6066.CIR-13-0220
Chang C-Y, Park H, Malone DC et al (2020) Immune checkpoint inhibitors and immune-related adverse events in patients with advanced melanoma. JAMA Netw Open 3:e201611. https://doi.org/10.1001/jamanetworkopen.2020.1611
Chaudhary K, Poirion OB, Lu L, Garmire LX (2018) Deep Learning based multi-omics integration robustly predicts survival in liver cancer. Clin Cancer Res 24:1248–1259. https://doi.org/10.1158/1078-0432.CCR-17-0853
Chen Y-Y, Chang W-A, Lin E-S et al (2019) Expressions of HLA class II genes in cutaneous melanoma were associated with clinical outcome: bioinformatics approaches and systematic analysis of public microarray and RNA-Seq datasets. Diagnostics (basel) 9:59. https://doi.org/10.3390/diagnostics9020059
Chen L, Wen Y, Xiong J et al (2023) An immunogenic cell death-related gene signature reflects immune landscape and predicts prognosis in melanoma independently of BRAF V600E status. Biomed Res Int 2023:1189022. https://doi.org/10.1155/2023/1189022
D’Arcy C, Kiel C (2021) Cell adhesion molecules in normal skin and melanoma. Biomolecules 11:1213. https://doi.org/10.3390/biom11081213
di Carlo E, de Totero D, Piazza T et al (2007) Role of IL-21 in immune-regulation and tumor immunotherapy. Cancer Immunol Immunother 56:1323–1334. https://doi.org/10.1007/s00262-007-0326-z
Gencia I, Baderca F, Avram S et al (2020) A preliminary study of microRNA expression in different types of primary melanoma. Bosn J Basic Med Sci 20:197–208. https://doi.org/10.17305/bjbms.2019.4271
Hao Y, Chen Y, He X et al (2020) Near-infrared responsive 5-fluorouracil and indocyanine green loaded MPEG-PCL nanoparticle integrated with dissolvable microneedle for skin cancer therapy. Bioact Mater 5:542–552. https://doi.org/10.1016/j.bioactmat.2020.04.002
Haymaker C, Johnson DH, Murthy R et al (2021) Tilsotolimod with ipilimumab drives tumor responses in anti-PD-1 refractory melanoma. Cancer Discov 11:1996. https://doi.org/10.1158/2159-8290.CD-20-1546
Ji Z-H, Ren W-Z, Yang S et al (2022) Identification of immune-related biomarkers associated with tumorigenesis and prognosis in skin cutaneous melanoma. Am J Cancer Res 12:1727–1739
Kramer MA (1991) Nonlinear principal component analysis using autoassociative neural networks. AIChE J 37:233–243. https://doi.org/10.1002/aic.690370209
Lê S, Josse J, Husson F (2008) FactoMineR: an R package for multivariate analysis. J Stat Softw 25:1–18. https://doi.org/10.18637/jss.v025.i01
Leng D, Zheng L, Wen Y et al (2022) A benchmark study of deep learning-based multi-omics data fusion methods for cancer. Genome Biol 23:171. https://doi.org/10.1186/s13059-022-02739-2
Longato E, Vettoretti M, Di Camillo B (2020) A practical perspective on the concordance index for the evaluation and selection of prognostic time-to-event models. J Biomed Inform 108:103496. https://doi.org/10.1016/j.jbi.2020.103496
Lunavat TR, Cheng L, Kim D-K et al (2015) Small RNA deep sequencing discriminates subsets of extracellular vesicles released by melanoma cells—evidence of unique microRNA cargos. RNA Biol 12:810–823. https://doi.org/10.1080/15476286.2015.1056975
Lv J, Wang J, Shang X et al (2020) Survival prediction in patients with colon adenocarcinoma via multiomics data integration using a deep learning algorithm. Biosci Rep. https://doi.org/10.1042/BSR20201482
Lv H, Liu X, Zeng X et al (2022) Comprehensive analysis of cuproptosis-related genes in immune infiltration and prognosis in melanoma. Front Pharmacol 13:930041. https://doi.org/10.3389/fphar.2022.930041
Marie KL, Sassano A, Yang HH et al (2020) Melanoblast transcriptome analysis reveals pathways promoting melanoma metastasis. Nat Commun 11:333. https://doi.org/10.1038/s41467-019-14085-2
Meyer D, Dimitriadou E, Hornik K, et al (2014) Misc Functions of the Department of Statistics (e1071), TU Wien
Miller D, Tallmadge RL, Binns M et al (2017) Polymorphism at expressed DQ and DR loci in five common equine MHC haplotypes. Immunogenetics 69:145–156. https://doi.org/10.1007/s00251-016-0964-4
Moin AT, Sarkar B, Ullah MA et al (2021) In silico assessment of EpCAM transcriptional expression and determination of the prognostic biomarker for human lung adenocarcinoma (LUAD) and lung squamous cell carcinoma (LUSC). Biochem Biophys Rep 27:101074. https://doi.org/10.1016/j.bbrep.2021.101074
Mondello P, Tadros S, Teater M et al (2020) Selective inhibition of HDAC3 targets synthetic vulnerabilities and activates immune surveillance in lymphoma. Cancer Discov 10:440–459. https://doi.org/10.1158/2159-8290.CD-19-0116
Monteleone G, Monteleone I, Fina D et al (2005) Interleukin-21 enhances T-helper cell type I signaling and interferon-γ production in Crohn’s disease. Gastroenterology 128:687–694. https://doi.org/10.1053/j.gastro.2004.12.042
Pavey S, Pinder A, Fernando W et al (2020) Multiple interaction nodes define the postreplication repair response to UV-induced DNA damage that is defective in melanomas and correlated with UV signature mutation load. Mol Oncol 14:22. https://doi.org/10.1002/1878-0261.12601
Peng J, Wang T, Yue C et al (2022) PGAM5: A necroptosis gene associated with poor tumor prognosis that promotes cutaneous melanoma progression. Front Oncol 12:1004511. https://doi.org/10.3389/fonc.2022.1004511
Petrella TM, Tozer R, Belanger K et al (2012) Interleukin-21 has activity in patients with metastatic melanoma: a phase II study. J Clin Oncol 30:3396–3401. https://doi.org/10.1200/JCO.2011.40.0655
Poirion OB, Chaudhary K, Garmire LX (2018) Deep Learning data integration for better risk stratification models of bladder cancer. AMIA Jt Summits Transl Sci Proc 2017:197–206
Ritchie ME, Phipson B, Wu D et al (2015) limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res 43:e47. https://doi.org/10.1093/nar/gkv007
Schadendorf D, Van Akkooi ACJ, Berking C et al (2018) Melanoma. Lancet 392:971–984. https://doi.org/10.1016/S0140-6736(18)31559-9
Schröder MS, Culhane AC, Quackenbush J, Haibe-Kains B (2011) survcomp: an R/Bioconductor package for performance assessment and comparison of survival models. Bioinformatics 27:3206–3208. https://doi.org/10.1093/bioinformatics/btr511
Shi Q, He Q, Wei J (2018) MicroRNA-342 prohibits proliferation and invasion of melanoma cells by directly targeting zinc-finger E-box-binding Homeobox 1. Oncol Res 26:1447–1455. https://doi.org/10.3727/096504018X15193823766141
Siegel RL, Miller KD, Jemal A (2017) Cancer Statistics, 2017. CA Cancer J Clin 67:10. https://doi.org/10.3322/caac.21387
Sun X, Zhang C, Cao Y, Liu E (2019) miR-150 suppresses tumor growth in melanoma through downregulation of MYB. Oncol Res 27:317–323. https://doi.org/10.3727/096504018X15228863026239
Tanagala KKK, Morin-Baxter J, Carvajal R et al (2022) SP140 inhibits STAT1 signaling, induces IFN-γ in tumor-associated macrophages, and is a predictive biomarker of immunotherapy response. J Immunother Cancer 10:e005088. https://doi.org/10.1136/jitc-2022-005088
Tay JK, Narasimhan B, Hastie T (2023) Elastic net regularization paths for all generalized linear models. J Stat Softw 106:1–31. https://doi.org/10.18637/jss.v106.i01
Team R (2014) R: A language and environment for statistical computing. MSOR connections
Tembe V, Schramm S-J, Stark MS et al (2015) MicroRNA and mRNA expression profiling in metastatic melanoma reveal associations with BRAF mutation and patient prognosis. Pigment Cell Melanoma Res 28:254–266. https://doi.org/10.1111/pcmr.12343
Tian Y, Morris TJ, Webster AP et al (2017) ChAMP: updated methylation analysis pipeline for Illumina BeadChips. Bioinformatics 33:3982–3984. https://doi.org/10.1093/bioinformatics/btx513
Wan Q, Wei R, Wei X, Deng Y (2022) Crosstalk of four kinds of cell deaths defines subtypes of cutaneous melanoma for precise immunotherapy and chemotherapy. Front Immunol 13:998454. https://doi.org/10.3389/fimmu.2022.998454
Wang T-H, Lee C-Y, Lee T-Y et al (2021) Biomarker identification through multiomics data analysis of prostate cancer prognostication using a deep learning model and similarity network fusion. Cancers (basel) 13:2528. https://doi.org/10.3390/cancers13112528
Wu T, Hu E, Xu S et al (2021) clusterProfiler 40: A universal enrichment tool for interpreting omics data. Innovation (camb) 2:100141. https://doi.org/10.1016/j.xinn.2021.100141
Xiong J, Bing Z, Guo S (2019) Observed survival interval: a supplement to TCGA pan-cancer clinical data resource. Cancers (basel) 11:280. https://doi.org/10.3390/cancers11030280
Xue D, Yang P, Wei Q et al (2019) IL-21/IL-21R inhibit tumor growth and invasion in non-small cell lung cancer cells via suppressing Wnt/β-catenin signaling and PD-L1 expression. Int J Mol Med 44:1697–1706. https://doi.org/10.3892/ijmm.2019.4354
Yang W, Jiang J, Schnellinger EM et al (2022) Modified Brier score for evaluating prediction accuracy for binary outcomes. Stat Methods Med Res 31:2287–2296. https://doi.org/10.1177/09622802221122391
Zhang X, Yan Z, Zhang J et al (2011) Combination of hsa-miR-375 and hsa-miR-142-5p as a predictor for recurrence risk in gastric cancer patients following surgical resection. Ann Oncol 22:2257–2266. https://doi.org/10.1093/annonc/mdq758
Zhang L, Lv C, Jin Y et al (2018) Deep learning-based multi-omics data integration reveals two prognostic subtypes in high-risk neuroblastoma. Front Genet 9:477. https://doi.org/10.3389/fgene.2018.00477
Zhang Y, Gao M, Chen L et al (2020) Licochalcone A restrains microphthalmia-associated transcription factor expression and growth by activating autophagy in melanoma cells via miR-142-3p/Rheb/mTOR pathway. Phytother Res 34:349–358. https://doi.org/10.1002/ptr.6525
Zhang C, Dang D, Cong L et al (2021a) Pivotal factors associated with the immunosuppressive tumor microenvironment and melanoma metastasis. Cancer Med 10:4710–4720. https://doi.org/10.1002/cam4.3963
Zhang E, Chen Y, Bao S et al (2021b) Identification of subgroups along the glycolysis-cholesterol synthesis axis and the development of an associated prognostic risk model. Hum Genomics 15:53. https://doi.org/10.1186/s40246-021-00350-3
Zhang X, Wang J, Lu J et al (2021c) Robust prognostic subtyping of muscle-invasive bladder cancer revealed by deep learning-based multi-omics data integration. Front Oncol 11:689626. https://doi.org/10.3389/fonc.2021.689626
Zhou X, Yan T, Huang C et al (2018) Melanoma cell-secreted exosomal miR-155-5p induce proangiogenic switch of cancer-associated fibroblasts via SOCS1/JAK2/STAT3 signaling pathway. J Exp Clin Cancer Res 37:242. https://doi.org/10.1186/s13046-018-0911-3
Acknowledgements
We would like to thank the TCGA database and GEO database platform, and the providers who uploaded SKCM datasets in the platform. Thanks are also due to all developers of the R packages covered in this paper.
Funding
This work was supported by the National Science Foundation of China (no. 81773312); Natural Science Foundation of Guangdong Province (no. 2015A030313517); Talents Recruitment Grant of “Yangfan Plan” of Guangdong Province (no. 201433005); and Social Development Science and Technology Project of Dongguan (no. 20211800905552).
Author information
Authors and Affiliations
Contributions
Conception and design: LWJ. Data collection and interpretation: LWJ, PY, HQ. Statistical analysis: LWJ and PY. Writing original draft: LWJ, HQ and PY. Writing review and editing: all authors. Supervision: HQ. Final approval of manuscript: all authors.
Corresponding author
Ethics declarations
Conflict of interest
The authors declare no conflict of interest.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary Information
Below is the link to the electronic supplementary material.
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Li, W., Huang, Q., Peng, Y. et al. A deep learning approach based on multi-omics data integration to construct a risk stratification prediction model for skin cutaneous melanoma. J Cancer Res Clin Oncol 149, 15923–15938 (2023). https://doi.org/10.1007/s00432-023-05358-x
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00432-023-05358-x