A Novel Hybrid Feature Selection and Ensemble Learning Framework for Unbalanced Cancer Data Diagnosis With Transcriptome and Functional Proteomic

Abstract

The high dimension, high redundancy and class imbalance of cancer multiple omics data are the main challenges for cancer diagnosis. Existing studies have neglected the role of functional proteomics in the occurrence and development of cancer. In this study, a novel hybrid feature selection and ensemble learning framework, referred to as the three-stage feature selection and twice-competitional ensemble learning method (TSFS-TCEM), is proposed for cancer diagnosis. Firstly, we combine the transcriptome and functional proteomics data to construct a multi-omics data on breast cancer, which is the first time to apply these combined biological data for diagnosing breast cancer. Secondly, the proposed method introduces multiple models during the feature selection and diagnostic model construction. The three-stage feature selections integrate the features from different types of data and the twice-competitional ensemble learning framework resolves the data imbalance problem suffer from a single classifier. The TSFS-TCEM achieves a diagnostic accuracy of 99.64%, outperforming all compared methods. In addition, the 5-fold cross-validation sensitivity, specificity and F-Measure of the method are above 99.63%.

Bias and Class Imbalance in Oncologic Data—Towards Inclusive and Transferrable AI in Large Scale Oncology Data Sets

“…Tang et al constructed multi-omics data with the transcriptome and functional proteomics data and employed a bagging-based ensemble learning method on breast cancer to avoid over-fitting problems [ 53 ]. Le et al utilized an ensemble of pre-trained ResNet50 CNN models for the classification of skin cancer to support dermatologists in skin cancer diagnosis [ 54 ].…”

“…Feature selection approaches are considered dimensionality reduction techniques. Due to the high dimensionality of proteomic data, analyzing, storing, training, and classifying these data may be considered an NP-Hard problem [ 53 , 60 ]. Feature selection methods to find optimal feature subsets reduce the computational and storing cost and eliminate redundant and irrelevant information, facilitating data visualization.…”

“…Their deep imaging feature-based biomarker provides a potential for the preoperative care of GBM patients [ 66 ]. Tang et al employed multi-omics data (transcriptome profiles and functional proteomics data) for breast cancer diagnosis [ 53 ], introducing a novel framework with hybrid feature selection and bagging-based ensemble learning methods to resolve the class imbalance problem. Wu et al demonstrated that dosiomics improves predicting locoregional recurrence for head and neck cancer cases and should be proposed correlatively.…”

A New Multi-Class Rebalancing Framework for Imbalance Medical Data

“…The framework performs significantly well in predicting colorectal cancer. Tang et al [9] proposed a hybrid framework with a combination of feature selection and ensemble-based learning called the Three-stage Feature selection and Twice-competitional Ensemble learning Method (TSFS-RCEM). This comprises of three-stage; the first stage is to perform information gain (IG) towards the imbalanced data, the second stage involves reducing its high dimensionality, and the final stage, includes feature selection to select the most relevant features.…”

KK-DBP: A Multi-Feature Fusion Method for DNA-Binding Protein Identification Based on Random Forest

“…Feature extraction methods can be broadly classified into two categories: those based on structural information and those based on sequence information ( Kim et al, 2004 ; Meng and Kurgan, 2016 ; Qu et al, 2019 ; Ao et al, 2021a ; Lv et al, 2021a ; Liu et al, 2021 ; Tang et al, 2021 ; Wu and Yu, 2021 ); ( Stawiski et al, 2003 ) proposed a model based on protein structure that utilises a neural network approach incorporating information like residue and hydrogen bond potential. Liu et al ( Liu et al, 2014 ) developed a model called IDNA-prot|dis, based on the pseudo amino acid composition (PseAAC) of protein sequence information.…”