Feature Selection in Cancer Classification: Utilizing Explainable Artificial Intelligence to Uncover Influential Genes in Machine Learning Models
This study investigates the use of machine learning (ML) models combined with explainable artificial intelligence (XAI) techniques to identify the most influential genes in the classification of five recurrent cancer types in women: breast cancer (BRCA), lung adenocarcinoma (LUAD), thyroid cancer (T...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2024-12-01
|
| Series: | AI |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2673-2688/6/1/2 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832589437380329472 |
|---|---|
| author | Matheus Dalmolin Karolayne S. Azevedo Luísa C. de Souza Caroline B. de Farias Martina Lichtenfels Marcelo A. C. Fernandes |
| author_facet | Matheus Dalmolin Karolayne S. Azevedo Luísa C. de Souza Caroline B. de Farias Martina Lichtenfels Marcelo A. C. Fernandes |
| author_sort | Matheus Dalmolin |
| collection | DOAJ |
| description | This study investigates the use of machine learning (ML) models combined with explainable artificial intelligence (XAI) techniques to identify the most influential genes in the classification of five recurrent cancer types in women: breast cancer (BRCA), lung adenocarcinoma (LUAD), thyroid cancer (THCA), ovarian cancer (OV), and colon adenocarcinoma (COAD). Gene expression data from RNA-seq, extracted from The Cancer Genome Atlas (TCGA), were used to train ML models, including decision trees (DTs), random forest (RF), and XGBoost (XGB), which achieved accuracies of 98.69%, 99.82%, and 99.37%, respectively. However, the challenges in this analysis included the high dimensionality of the dataset and the lack of transparency in the ML models. To mitigate these challenges, the SHAP (Shapley Additive Explanations) method was applied to generate a list of features, aiming to understand which characteristics influenced the models’ decision-making processes and, consequently, the prediction results for the five tumor types. The SHAP analysis identified 119, 80, and 10 genes for the RF, XGB, and DT models, respectively, totaling 209 genes, resulting in 172 unique genes. The new list, representing 0.8% of the original input features, is coherent and fully explainable, increasing confidence in the applied models. Additionally, the results suggest that the SHAP method can be effectively used as a feature selector in gene expression data. This approach not only enhances model transparency but also maintains high classification performance, highlighting its potential in identifying biologically relevant features that may serve as biomarkers for cancer diagnostics and treatment planning. |
| format | Article |
| id | doaj-art-e198b00f85d748838c7451a946da554a |
| institution | Kabale University |
| issn | 2673-2688 |
| language | English |
| publishDate | 2024-12-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | AI |
| spelling | doaj-art-e198b00f85d748838c7451a946da554a2025-01-24T13:17:21ZengMDPI AGAI2673-26882024-12-0161210.3390/ai6010002Feature Selection in Cancer Classification: Utilizing Explainable Artificial Intelligence to Uncover Influential Genes in Machine Learning ModelsMatheus Dalmolin0Karolayne S. Azevedo1Luísa C. de Souza2Caroline B. de Farias3Martina Lichtenfels4Marcelo A. C. Fernandes5InovAI Lab, Federal University of Rio Grande do Norte, Natal 59078-970, RN, BrazilInovAI Lab, Federal University of Rio Grande do Norte, Natal 59078-970, RN, BrazilInovAI Lab, Federal University of Rio Grande do Norte, Natal 59078-970, RN, BrazilNational Science and Technology Institute for Children’s Cancer Biology and Pediatric Oncology-INCT BioOncoPed, Porto Alegre 90620-110, RS, BrazilZiel Biosciences, Porto Alegre 90650-001, RS, BrazilInovAI Lab, Federal University of Rio Grande do Norte, Natal 59078-970, RN, BrazilThis study investigates the use of machine learning (ML) models combined with explainable artificial intelligence (XAI) techniques to identify the most influential genes in the classification of five recurrent cancer types in women: breast cancer (BRCA), lung adenocarcinoma (LUAD), thyroid cancer (THCA), ovarian cancer (OV), and colon adenocarcinoma (COAD). Gene expression data from RNA-seq, extracted from The Cancer Genome Atlas (TCGA), were used to train ML models, including decision trees (DTs), random forest (RF), and XGBoost (XGB), which achieved accuracies of 98.69%, 99.82%, and 99.37%, respectively. However, the challenges in this analysis included the high dimensionality of the dataset and the lack of transparency in the ML models. To mitigate these challenges, the SHAP (Shapley Additive Explanations) method was applied to generate a list of features, aiming to understand which characteristics influenced the models’ decision-making processes and, consequently, the prediction results for the five tumor types. The SHAP analysis identified 119, 80, and 10 genes for the RF, XGB, and DT models, respectively, totaling 209 genes, resulting in 172 unique genes. The new list, representing 0.8% of the original input features, is coherent and fully explainable, increasing confidence in the applied models. Additionally, the results suggest that the SHAP method can be effectively used as a feature selector in gene expression data. This approach not only enhances model transparency but also maintains high classification performance, highlighting its potential in identifying biologically relevant features that may serve as biomarkers for cancer diagnostics and treatment planning.https://www.mdpi.com/2673-2688/6/1/2explainable AImachine learningfeature selectionRNA-seqcancerSHAP |
| spellingShingle | Matheus Dalmolin Karolayne S. Azevedo Luísa C. de Souza Caroline B. de Farias Martina Lichtenfels Marcelo A. C. Fernandes Feature Selection in Cancer Classification: Utilizing Explainable Artificial Intelligence to Uncover Influential Genes in Machine Learning Models AI explainable AI machine learning feature selection RNA-seq cancer SHAP |
| title | Feature Selection in Cancer Classification: Utilizing Explainable Artificial Intelligence to Uncover Influential Genes in Machine Learning Models |
| title_full | Feature Selection in Cancer Classification: Utilizing Explainable Artificial Intelligence to Uncover Influential Genes in Machine Learning Models |
| title_fullStr | Feature Selection in Cancer Classification: Utilizing Explainable Artificial Intelligence to Uncover Influential Genes in Machine Learning Models |
| title_full_unstemmed | Feature Selection in Cancer Classification: Utilizing Explainable Artificial Intelligence to Uncover Influential Genes in Machine Learning Models |
| title_short | Feature Selection in Cancer Classification: Utilizing Explainable Artificial Intelligence to Uncover Influential Genes in Machine Learning Models |
| title_sort | feature selection in cancer classification utilizing explainable artificial intelligence to uncover influential genes in machine learning models |
| topic | explainable AI machine learning feature selection RNA-seq cancer SHAP |
| url | https://www.mdpi.com/2673-2688/6/1/2 |
| work_keys_str_mv | AT matheusdalmolin featureselectionincancerclassificationutilizingexplainableartificialintelligencetouncoverinfluentialgenesinmachinelearningmodels AT karolaynesazevedo featureselectionincancerclassificationutilizingexplainableartificialintelligencetouncoverinfluentialgenesinmachinelearningmodels AT luisacdesouza featureselectionincancerclassificationutilizingexplainableartificialintelligencetouncoverinfluentialgenesinmachinelearningmodels AT carolinebdefarias featureselectionincancerclassificationutilizingexplainableartificialintelligencetouncoverinfluentialgenesinmachinelearningmodels AT martinalichtenfels featureselectionincancerclassificationutilizingexplainableartificialintelligencetouncoverinfluentialgenesinmachinelearningmodels AT marceloacfernandes featureselectionincancerclassificationutilizingexplainableartificialintelligencetouncoverinfluentialgenesinmachinelearningmodels |