1
|
Mirza B, Wang W, Wang J, Choi H, Chung NC, Ping P. Machine Learning and Integrative Analysis of Biomedical Big Data. Genes (Basel) 2019; 10:E87. [PMID: 30696086 PMCID: PMC6410075 DOI: 10.3390/genes10020087] [Citation(s) in RCA: 157] [Impact Index Per Article: 31.4] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 12/02/2018] [Revised: 01/08/2019] [Accepted: 01/21/2019] [Indexed: 12/11/2022] Open
Abstract
Recent developments in high-throughput technologies have accelerated the accumulation of massive amounts of omics data from multiple sources: genome, epigenome, transcriptome, proteome, metabolome, etc. Traditionally, data from each source (e.g., genome) is analyzed in isolation using statistical and machine learning (ML) methods. Integrative analysis of multi-omics and clinical data is key to new biomedical discoveries and advancements in precision medicine. However, data integration poses new computational challenges as well as exacerbates the ones associated with single-omics studies. Specialized computational approaches are required to effectively and efficiently perform integrative analysis of biomedical data acquired from diverse modalities. In this review, we discuss state-of-the-art ML-based approaches for tackling five specific computational challenges associated with integrative analysis: curse of dimensionality, data heterogeneity, missing data, class imbalance and scalability issues.
Collapse
Affiliation(s)
- Bilal Mirza
- NIH BD2K Center of Excellence for Biomedical Computing, University of California Los Angeles, Los Angeles, CA 90095, USA.
- Department of Physiology, University of California Los Angeles, Los Angeles, CA 90095, USA.
| | - Wei Wang
- NIH BD2K Center of Excellence for Biomedical Computing, University of California Los Angeles, Los Angeles, CA 90095, USA.
- Department of Computer Science, University of California Los Angeles, Los Angeles, CA 90095, USA.
- Scalable Analytics Institute (ScAi), University of California Los Angeles, Los Angeles, CA 90095, USA.
- Department of Bioinformatics, University of California Los Angeles, Los Angeles, CA 90095, USA.
| | - Jie Wang
- NIH BD2K Center of Excellence for Biomedical Computing, University of California Los Angeles, Los Angeles, CA 90095, USA.
- Department of Physiology, University of California Los Angeles, Los Angeles, CA 90095, USA.
| | - Howard Choi
- NIH BD2K Center of Excellence for Biomedical Computing, University of California Los Angeles, Los Angeles, CA 90095, USA.
- Department of Physiology, University of California Los Angeles, Los Angeles, CA 90095, USA.
- Department of Bioinformatics, University of California Los Angeles, Los Angeles, CA 90095, USA.
| | - Neo Christopher Chung
- NIH BD2K Center of Excellence for Biomedical Computing, University of California Los Angeles, Los Angeles, CA 90095, USA.
- Institute of Informatics, Faculty of Mathematics, Informatics and Mechanics, University of Warsaw, Banacha 2, 02-097 Warsaw, Poland.
| | - Peipei Ping
- NIH BD2K Center of Excellence for Biomedical Computing, University of California Los Angeles, Los Angeles, CA 90095, USA.
- Department of Physiology, University of California Los Angeles, Los Angeles, CA 90095, USA.
- Scalable Analytics Institute (ScAi), University of California Los Angeles, Los Angeles, CA 90095, USA.
- Department of Bioinformatics, University of California Los Angeles, Los Angeles, CA 90095, USA.
- Department of Medicine (Cardiology), University of California Los Angeles, Los Angeles, CA 90095, USA.
| |
Collapse
|
2
|
Yu S, Zhao H. Rough sets and Laplacian score based cost-sensitive feature selection. PLoS One 2018; 13:e0197564. [PMID: 29912884 PMCID: PMC6005488 DOI: 10.1371/journal.pone.0197564] [Citation(s) in RCA: 3] [Impact Index Per Article: 0.5] [Reference Citation Analysis] [Abstract] [MESH Headings] [Grants] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 09/24/2017] [Accepted: 12/10/2017] [Indexed: 12/02/2022] Open
Abstract
Cost-sensitive feature selection learning is an important preprocessing step in machine learning and data mining. Recently, most existing cost-sensitive feature selection algorithms are heuristic algorithms, which evaluate the importance of each feature individually and select features one by one. Obviously, these algorithms do not consider the relationship among features. In this paper, we propose a new algorithm for minimal cost feature selection called the rough sets and Laplacian score based cost-sensitive feature selection. The importance of each feature is evaluated by both rough sets and Laplacian score. Compared with heuristic algorithms, the proposed algorithm takes into consideration the relationship among features with locality preservation of Laplacian score. We select a feature subset with maximal feature importance and minimal cost when cost is undertaken in parallel, where the cost is given by three different distributions to simulate different applications. Different from existing cost-sensitive feature selection algorithms, our algorithm simultaneously selects out a predetermined number of “good” features. Extensive experimental results show that the approach is efficient and able to effectively obtain the minimum cost subset. In addition, the results of our method are more promising than the results of other cost-sensitive feature selection algorithms.
Collapse
Affiliation(s)
- Shenglong Yu
- Fujian Key Laboratory of Granular Computing and Application (Minnan Normal University), Zhangzhou, Fujian, China
- Key Laboratory of Data Science and Intelligence Application, Fujian Province University, Zhangzhou, Fujian, China
| | - Hong Zhao
- Fujian Key Laboratory of Granular Computing and Application (Minnan Normal University), Zhangzhou, Fujian, China
- Key Laboratory of Data Science and Intelligence Application, Fujian Province University, Zhangzhou, Fujian, China
- * E-mail:
| |
Collapse
|
3
|
Pardini B, Cordero F, Naccarati A, Viberti C, Birolo G, Oderda M, Di Gaetano C, Arigoni M, Martina F, Calogero RA, Sacerdote C, Gontero P, Vineis P, Matullo G. microRNA profiles in urine by next-generation sequencing can stratify bladder cancer subtypes. Oncotarget 2018; 9:20658-20669. [PMID: 29755679 PMCID: PMC5945522 DOI: 10.18632/oncotarget.25057] [Citation(s) in RCA: 46] [Impact Index Per Article: 7.7] [Reference Citation Analysis] [Abstract] [Key Words] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 09/06/2017] [Accepted: 03/18/2018] [Indexed: 12/31/2022] Open
Abstract
Bladder cancer (BC) is the most frequent malignancy of the urinary tract with a high incidence in men and smokers. Currently, there are no non-invasive markers useful for BC diagnosis and subtypes classification that could overcome invasive procedures such as cystoscopy. Dysregulated miRNA profiles have been associated with numerous cancers, including BC. Cell-free miRNAs are abundantly present in a variety of biofluids including urine and make them promising candidates in cancer biomarker discovery. In the present study, the identification of miRNA fingerprints associated with different BC status was performed by next-generation sequencing on urine samples from 66 BC and 48 controls. Three signatures based on dysregulated miRNAs have been identified by regression models, assessing the power to discriminate different BC subtypes. Altered miRNAs according to invasiveness and grade were validated by qPCR on 112 cases and 65 controls (among which 46 cases and 16 controls were an independent group of subjects while the rest were replica samples). The area under the curve (AUC) computed including three miRNAs (miR-30a-5p, let-7c-5p and miR-486-5p) altered in all BC subtypes showed a significantly increased accuracy in the discrimination of cases and controls (AUC model = 0.70; p-value = 0.01). In conclusions, the non-invasive detection in urine of a selected number of miRNAs altered in different BC subtypes could lead to an accurate early diagnosis of cancer and stratification of patients.
Collapse
Affiliation(s)
- Barbara Pardini
- Italian Institute for Genomic Medicine, Turin, Italy.,Department of Medical Sciences, University of Turin, Turin, Italy
| | | | | | - Clara Viberti
- Italian Institute for Genomic Medicine, Turin, Italy.,Department of Medical Sciences, University of Turin, Turin, Italy
| | - Giovanni Birolo
- Italian Institute for Genomic Medicine, Turin, Italy.,Department of Medical Sciences, University of Turin, Turin, Italy
| | - Marco Oderda
- Department of Surgical Sciences, University of Turin and Città della Salute e della Scienza, Turin, Italy
| | - Cornelia Di Gaetano
- Italian Institute for Genomic Medicine, Turin, Italy.,Department of Medical Sciences, University of Turin, Turin, Italy
| | - Maddalena Arigoni
- Molecular Biotechnology Center, Department of Biotechnology and Health Sciences, University of Turin, Turin, Italy
| | - Federica Martina
- Department of Computer Science, University of Turin, Turin, Italy
| | - Raffaele A Calogero
- Molecular Biotechnology Center, Department of Biotechnology and Health Sciences, University of Turin, Turin, Italy
| | | | - Paolo Gontero
- Department of Surgical Sciences, University of Turin and Città della Salute e della Scienza, Turin, Italy
| | - Paolo Vineis
- Italian Institute for Genomic Medicine, Turin, Italy.,MRC-HPA Centre for Environment and Health, School of Public Health, Imperial College London, London, United Kingdom
| | - Giuseppe Matullo
- Italian Institute for Genomic Medicine, Turin, Italy.,Department of Medical Sciences, University of Turin, Turin, Italy
| |
Collapse
|