신장암의 임상 예측을 위한 딥 오토인코더 기반 분류 = Deep Autoencoder based Classification for Clinical Prediction of Kidney Cancer

다국어 초록 (Multilingual Abstract)

Predicting clinical information using gene expression is challenging given the complexity and high dimensionality of gene data. This study propose a deep learning framework for cancer diagnosis through feature extraction and classifier based on various pre-trained autoencoder technologies for kidney cancer. It can be fine-tuned for any tasks and predict clinical information by neural network classifiers. Our model achieved micro and macro F1-scores of 96.2% and 95.8% for gender, 95.8% and 76.3% for race, and 99.8% and 99.6% for sample type predictions, respectively, which is much higher than the values of traditional dimensionality reduction and machine learning techniques. In the results, the conditional variational mutation autoencoder (CVAE) improved the macro F1 score, a difficult race prediction task, by 7.6%. Our results are useful for the prognosis as well as prevention and early diagnosis of kidney cancer.

참고문헌 (Reference)

1 지병훈 ; 장인호, "한국인에서 신장암의 과잉 진단 및 작은 국소신장암에서 능동적 관찰의 의미" 대한비뇨기종양학회 16 (16): 15-24, 2018

2 K. Simonyan, "Very deep convolutional networks for large-scale image recognition" 2015

3 Y. Pu, "Variational autoencoder for deep learning of images, labels and captions" 2016

4 Nikola Simidjievski, "Variational Autoencoders for Cancer Data Integration: Design Principles and Computational Practice" Frontiers Media SA 10 : 2019

5 W. M, Linehan, "The genetic basis of cancer of the kidney" 170 : 2163-2172, 2003

6 Y. Zhan, "Systematic Analysis of the Global, Regional and National Burden of Kidney Cancer from 1990 to 2017 : Results from the Global Burden of Disease Study 2017" 8 (8): 302-319, 2021

7 L. Le, "Supervised autoencoders : Improving generalization performance with unsupervised regularizers" 2018

8 M. Amgad, "Structured crowdsourcing enables convolutional segmentation of histology images" 35 (35): 3461-3467, 2019

9 P. Vincent, "Stacked denoising autoencoders : Learning useful representations in a deep network with a local denoising criterion" 11 : 3371-3408, 2010

10 A. J. Peired, "Sex and Gender Differences in Kidney Cancer : Clinical and Experimental Evidence" 13 (13): 4588-, 2021

11 F. Pedregosa, "Scikit-learn : Machine learning in Python" 12 : 2825-2830, 2011

12 D. Hepps, "Risk of renal insufficiency in African-Americans after radical nephrectomy for kidney cancer" 24 (24): 391-395, 2006

13 J. J. Hsieh, "Renal cell carcinoma" 3 (3): 1-19, 2017

14 L. Lipworth, "Renal cell cancer histological subtype distribution differs by race and sex" 117 (117): 260-265, 2016

15 D. A. Siegel, "Rates and trends of pediatric acute lymphoblastic leukemia—United States, 2001–2014" 66 (66): 950-954, 2017

16 A. F. Olshan, "Racial difference in histologic subtype of renal cell carcinoma" 2 (2): 744-749, 2013

17 H. Y. Xiong, "RNA splicing. The human splicing code reveals new insights into the genetic determinants of disease" 347 (347): 1-20, 2015

18 A. Paszke, "Pytorch : An imperative style, high-performance deep learning library" 8026-8037, 2019

19 T. R. Rebbeck, "Prostate cancer disparities by race and ethnicity : from nucleotide to neighborhood" 8 (8): a030387-, 2018

20 B. J. Kim, "Prediction of inherited genomic susceptibility to 20 common cancer types by a supervised machine-learning method" 115 (115): 1322-1327, 2018

21 J. R. Vasselli, "Predicting survival in patients with metastatic kidney cancer by geneexpression profiling in the primary tumor" 100 (100): 6958-6963, 2003

22 P. Mamoshina, "Population specific biomarkers of human aging : a big data study using South Korean, Canadian, and Eastern European patient populations" 73 (73): 1482-1490, 2018

23 L. A. Torre, "Ovarian cancer statistics, 2018" 68 : 284-296, 2018

24 P. Baldi, "Neural networks and principal component analysis : Learning from examples without local minima" 2 : 53-58, 1989

25 "National Cancer Center"

26 H. M. Kim, "Machine learning approach to predict the probability of recurrence of renal cell carcinoma after surgery : Prediction model development study" 9 (9): e25635-, 2021

27 K. Sohn, "Learning structured output representation using deep conditional generative models" 2 : 3483-3491, 2015

28 N. Hadjiyski, "Kidney cancer staging: Deep learning neural network based approach" 2020

29 N. Chowdhury, "Kidney cancer : an overview of current therapeutic approaches" 47 (47): 419-431, 2020

30 V. M. G. Olivares, "Immunohistochemical profile of renal cell tumours" 52 (52): 214-221, 2019

31 V. M. G. Olivares, "Immunohistochemical profile of renal cell tumours" 52 (52): 214-221, 2019

32 "Genomic Data Commons"

33 M. Mohri, "Generalization bounds for supervised dimensionality reduction" 44 : 226-241, 2015

34 M. Mancini, "Gender-related approach to kidney cancer management : Moving forward" 21 (21): 3378-, 2020

35 I. Lucca, "Gender differences in incidence and outcomes of urothelial and kidney cancer" 12 (12): 585-592, 2015

36 M. Mohri, "Foundations of Machine Learning" MIT Press 2012

37 W. M. Linehan, "Focus on kidney cancer" 6 (6): 223-228, 2004

38 M. A. Ranzato, "Efficient learning of sparse representations with an energy-based model" 19 : 1137-1144, 2007

39 S. J. O. Nomura, "Dietary intake of soy and cruciferous vegetables and treatmentrelated symptoms in Chinese-American and non-Hispanic White breast cancer survivors" 168 (168): 467-479, 2017

40 B. Shuch, "Defining early-onset kidney cancer : implications for germline and somatic mutation testing and clinical management" 32 (32): 431-437, 2014

41 S. Belharbi, "Deep neural networks regularization for structured output prediction" 281 : 169-177, 2018

42 Y. Bengio, "Deep generative stochastic networks trainable by backprop" 32 : 226-234, 2014

43 M. Mostavi, "Convolutional neural network models for cancer type prediction based on gene expression" 13 (13): 1-13, 2020

44 Cancer Genome Atlas Research Network, "Comprehensive molecular characterization of papillary renal-cell carcinoma" 374 (374): 135-145, 2016

45 Ho Sun Shon, "Classification of Kidney Cancer Data Using Cost-Sensitive Hybrid Deep Learning Approach" MDPI AG 12 (12): 154-, 2020

46 D. P. Kingma, "Auto-encoding variational bayes" 2014

47 N. E. M. Khalifa, "Artificial intelligence technique for gene expression by tumor RNA-Seq data : a novel optimized deep learning approach" 8 : 22874-22883, 2020

48 L. A. Gottlieb, "Adaptive metric dimensionality reduction" 620 : 105-118, 2016

49 A. M. Ali, "A machine learning approach for the classification of kidney cancer subtypes using miRNA genome data" 8 (8): 1-14, 2018

50 R. Tabares-Soto, "A comparative study of machine learning and deep learning algorithms to classify cancer types based on microarray gene expression data" 6 : e270-, 2020

51 O. G. Troyanskaya, "A Bayesian framework for combining heterogeneous data sources for gene function prediction(in Saccharomyces cerevisiae)" 100 (100): 8348-8353, 2003