A robust ensemble feature selection approach to prioritize genes associated with survival outcome in high-dimensional gene expression data
Exploring features associated with the clinical outcome of interest is a rapidly advancing area of research. However, with contemporary sequencing technologies capable of identifying over thousands of genes per sample, there is a challenge in constructing efficient prediction models that balance acc...
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Frontiers Media S.A.
2024-03-01
|
| Series: | Frontiers in Systems Biology |
| Subjects: | |
| Online Access: | https://www.frontiersin.org/articles/10.3389/fsysb.2024.1355595/full |
| _version_ | 1797253990432374784 |
|---|---|
| author | Phi Le Xingyue Gong Leah Ung Hai Yang Bridget P. Keenan Bridget P. Keenan Li Zhang Li Zhang Li Zhang Tao He |
| author_facet | Phi Le Xingyue Gong Leah Ung Hai Yang Bridget P. Keenan Bridget P. Keenan Li Zhang Li Zhang Li Zhang Tao He |
| author_sort | Phi Le |
| collection | DOAJ |
| description | Exploring features associated with the clinical outcome of interest is a rapidly advancing area of research. However, with contemporary sequencing technologies capable of identifying over thousands of genes per sample, there is a challenge in constructing efficient prediction models that balance accuracy and resource utilization. To address this challenge, researchers have developed feature selection methods to enhance performance, reduce overfitting, and ensure resource efficiency. However, applying feature selection models to survival analysis, particularly in clinical datasets characterized by substantial censoring and limited sample sizes, introduces unique challenges. We propose a robust ensemble feature selection approach integrated with group Lasso to identify compelling features and evaluate its performance in predicting survival outcomes. Our approach consistently outperforms established models across various criteria through extensive simulations, demonstrating low false discovery rates, high sensitivity, and high stability. Furthermore, we applied the approach to a colorectal cancer dataset from The Cancer Genome Atlas, showcasing its effectiveness by generating a composite score based on the selected genes to correctly distinguish different subtypes of the patients. In summary, our proposed approach excels in selecting impactful features from high-dimensional data, yielding better outcomes compared to contemporary state-of-the-art models. |
| first_indexed | 2024-04-24T21:42:50Z |
| format | Article |
| id | doaj.art-a845c2977dcf43798780ba07759f8b5f |
| institution | Directory Open Access Journal |
| issn | 2674-0702 |
| language | English |
| last_indexed | 2024-04-24T21:42:50Z |
| publishDate | 2024-03-01 |
| publisher | Frontiers Media S.A. |
| record_format | Article |
| series | Frontiers in Systems Biology |
| spelling | doaj.art-a845c2977dcf43798780ba07759f8b5f2024-03-21T05:14:00ZengFrontiers Media S.A.Frontiers in Systems Biology2674-07022024-03-01410.3389/fsysb.2024.13555951355595A robust ensemble feature selection approach to prioritize genes associated with survival outcome in high-dimensional gene expression dataPhi Le0Xingyue Gong1Leah Ung2Hai Yang3Bridget P. Keenan4Bridget P. Keenan5Li Zhang6Li Zhang7Li Zhang8Tao He9Division of Hematology/Oncology, Department of Medicine, University of California, San Francisco, San Francisco, CA, United StatesDepartment of Physiological Nursing, School of Nursing, University of California, San Francisco, San Francisco, CA, United StatesDivision of Hematology/Oncology, Department of Medicine, University of California, San Francisco, San Francisco, CA, United StatesDivision of Hematology/Oncology, Department of Medicine, University of California, San Francisco, San Francisco, CA, United StatesDivision of Hematology/Oncology, Department of Medicine, University of California, San Francisco, San Francisco, CA, United StatesHelen Diller Family Comprehensive Cancer Center, University of California, San Francisco, San Francisco, CA, United StatesDivision of Hematology/Oncology, Department of Medicine, University of California, San Francisco, San Francisco, CA, United StatesHelen Diller Family Comprehensive Cancer Center, University of California, San Francisco, San Francisco, CA, United StatesDepartment of Epidemiology and Biostatistics, University of California, San Francisco, San Francisco, CA, United StatesDepartment of Mathematics, San Francisco State University, San Francisco, CA, United StatesExploring features associated with the clinical outcome of interest is a rapidly advancing area of research. However, with contemporary sequencing technologies capable of identifying over thousands of genes per sample, there is a challenge in constructing efficient prediction models that balance accuracy and resource utilization. To address this challenge, researchers have developed feature selection methods to enhance performance, reduce overfitting, and ensure resource efficiency. However, applying feature selection models to survival analysis, particularly in clinical datasets characterized by substantial censoring and limited sample sizes, introduces unique challenges. We propose a robust ensemble feature selection approach integrated with group Lasso to identify compelling features and evaluate its performance in predicting survival outcomes. Our approach consistently outperforms established models across various criteria through extensive simulations, demonstrating low false discovery rates, high sensitivity, and high stability. Furthermore, we applied the approach to a colorectal cancer dataset from The Cancer Genome Atlas, showcasing its effectiveness by generating a composite score based on the selected genes to correctly distinguish different subtypes of the patients. In summary, our proposed approach excels in selecting impactful features from high-dimensional data, yielding better outcomes compared to contemporary state-of-the-art models.https://www.frontiersin.org/articles/10.3389/fsysb.2024.1355595/fullcolorectal cancerensemble feature selectionhigh-dimensional datatime-to-event outcomepseudo variablesgroup lasso |
| spellingShingle | Phi Le Xingyue Gong Leah Ung Hai Yang Bridget P. Keenan Bridget P. Keenan Li Zhang Li Zhang Li Zhang Tao He A robust ensemble feature selection approach to prioritize genes associated with survival outcome in high-dimensional gene expression data Frontiers in Systems Biology colorectal cancer ensemble feature selection high-dimensional data time-to-event outcome pseudo variables group lasso |
| title | A robust ensemble feature selection approach to prioritize genes associated with survival outcome in high-dimensional gene expression data |
| title_full | A robust ensemble feature selection approach to prioritize genes associated with survival outcome in high-dimensional gene expression data |
| title_fullStr | A robust ensemble feature selection approach to prioritize genes associated with survival outcome in high-dimensional gene expression data |
| title_full_unstemmed | A robust ensemble feature selection approach to prioritize genes associated with survival outcome in high-dimensional gene expression data |
| title_short | A robust ensemble feature selection approach to prioritize genes associated with survival outcome in high-dimensional gene expression data |
| title_sort | robust ensemble feature selection approach to prioritize genes associated with survival outcome in high dimensional gene expression data |
| topic | colorectal cancer ensemble feature selection high-dimensional data time-to-event outcome pseudo variables group lasso |
| url | https://www.frontiersin.org/articles/10.3389/fsysb.2024.1355595/full |
| work_keys_str_mv | AT phile arobustensemblefeatureselectionapproachtoprioritizegenesassociatedwithsurvivaloutcomeinhighdimensionalgeneexpressiondata AT xingyuegong arobustensemblefeatureselectionapproachtoprioritizegenesassociatedwithsurvivaloutcomeinhighdimensionalgeneexpressiondata AT leahung arobustensemblefeatureselectionapproachtoprioritizegenesassociatedwithsurvivaloutcomeinhighdimensionalgeneexpressiondata AT haiyang arobustensemblefeatureselectionapproachtoprioritizegenesassociatedwithsurvivaloutcomeinhighdimensionalgeneexpressiondata AT bridgetpkeenan arobustensemblefeatureselectionapproachtoprioritizegenesassociatedwithsurvivaloutcomeinhighdimensionalgeneexpressiondata AT bridgetpkeenan arobustensemblefeatureselectionapproachtoprioritizegenesassociatedwithsurvivaloutcomeinhighdimensionalgeneexpressiondata AT lizhang arobustensemblefeatureselectionapproachtoprioritizegenesassociatedwithsurvivaloutcomeinhighdimensionalgeneexpressiondata AT lizhang arobustensemblefeatureselectionapproachtoprioritizegenesassociatedwithsurvivaloutcomeinhighdimensionalgeneexpressiondata AT lizhang arobustensemblefeatureselectionapproachtoprioritizegenesassociatedwithsurvivaloutcomeinhighdimensionalgeneexpressiondata AT taohe arobustensemblefeatureselectionapproachtoprioritizegenesassociatedwithsurvivaloutcomeinhighdimensionalgeneexpressiondata AT phile robustensemblefeatureselectionapproachtoprioritizegenesassociatedwithsurvivaloutcomeinhighdimensionalgeneexpressiondata AT xingyuegong robustensemblefeatureselectionapproachtoprioritizegenesassociatedwithsurvivaloutcomeinhighdimensionalgeneexpressiondata AT leahung robustensemblefeatureselectionapproachtoprioritizegenesassociatedwithsurvivaloutcomeinhighdimensionalgeneexpressiondata AT haiyang robustensemblefeatureselectionapproachtoprioritizegenesassociatedwithsurvivaloutcomeinhighdimensionalgeneexpressiondata AT bridgetpkeenan robustensemblefeatureselectionapproachtoprioritizegenesassociatedwithsurvivaloutcomeinhighdimensionalgeneexpressiondata AT bridgetpkeenan robustensemblefeatureselectionapproachtoprioritizegenesassociatedwithsurvivaloutcomeinhighdimensionalgeneexpressiondata AT lizhang robustensemblefeatureselectionapproachtoprioritizegenesassociatedwithsurvivaloutcomeinhighdimensionalgeneexpressiondata AT lizhang robustensemblefeatureselectionapproachtoprioritizegenesassociatedwithsurvivaloutcomeinhighdimensionalgeneexpressiondata AT lizhang robustensemblefeatureselectionapproachtoprioritizegenesassociatedwithsurvivaloutcomeinhighdimensionalgeneexpressiondata AT taohe robustensemblefeatureselectionapproachtoprioritizegenesassociatedwithsurvivaloutcomeinhighdimensionalgeneexpressiondata |