Performance Comparison of Deep Learning Autoencoders for Cancer Subtype Detection Using Multi-Omics Data

Abstract

A heterogeneous disease like cancer is activated through multiple pathways and different perturbations. Depending upon the activated pathway(s), patients’ survival vary significantly and show different efficacy to various drugs. Therefore, cancer subtype detection using genomics level data is a significant research problem. Subtype detection is often a complex problem, and in most cases, needs multi-omics data fusion to achieve accurate subtyping. Different data fusion and subtyping approaches have been proposed, such as kernel-based fusion, matrix factorization, and deep learning autoencoders. In this paper, we compared the performance of different deep learning autoencoders for cancer subtype detection. We performed cancer subtype detection on four different cancer types from The Cancer Genome Atlas (TCGA) datasets using four autoencoder implementations. We also predicted the optimal number of subtypes in a cancer type using the silhouette score. We observed that the detected subtypes exhibit significant differences in survival profiles. Furthermore, we also compared the effect of feature selection and similarity measures for subtype detection. To evaluate the results obtained, we selected the Glioblastoma multiforme (GBM) dataset and identified the differentially expressed genes in each of the subtypes identified by the autoencoders; the obtained results coincide well with other genomic studies and can be corroborated with the involved pathways and biological functions. Thus, it shows that the results from the autoencoders, obtained through the interaction of different datatypes of cancer, can be used for the prediction and characterization of patient subgroups and survival profiles.

Multimodal deep learning for biomedical data fusion: a review

“…Franco et al . [ 33 ] compared several AE types for early fusion on cancer survival subtyping with multi-omics data. Though regular AE and VAE architectures seemed to outperform other AEs, the strong variation between performance on different data sets indicate the importance of choice of architecture.…”

A benchmark study of deep learning-based multi-omics data fusion methods for cancer

“…However, multi-omics data are complex, high-dimensional, and heterogeneous [8,9], and it is challenging to extract valuable knowledge from these multi-omics data. To address this challenge, various methods have been developed, such as multiple kernel learning, Bayesian consensus clustering, machine learning (ML)-based dimensionality reduction, similarity network fusion, and deep learning (DL) methods [10,11].…”

A systematic literature review for the prediction of anticancer drug response using various machine‐learning and deep‐learning techniques

“…The year‐wise distribution of the total selected papers can be seen in the graph below in Figure 2, which mainly emphasizes predicting the anticancer drug response (Adam et al, 2020; Ahmadi Moughari & Eslahchi, 2020; Choi et al, 2020; Koch et al, 2020; Kong et al, 2020; Kurilov et al, 2020; Patel et al, 2020; Sharma & Rani, 2020a; Wang, Li, Carpenter, & Guan, 2020; Zhu et al, 2020). The years between 2020 and 2021 (Cilluffo et al, 2021; Feng et al, 2021; Franco et al, 2021; Gerdes et al, 2021; Kim et al, 2021; Lv et al, 2021; Mudali et al, 2020; Nguyen et al, 2021; Partin et al, 2021; Patel & Shah, 2021; Piroozmand et al, 2020; Rafique et al, 2021; Schperberg et al, 2020; Vatansever et al, 2021; Vougas et al, 2020; Wang, Li, & Guan, 2020; Yu et al, 2021; Zhang et al, 2021) recorded the maximum number of publications up to 21–28 articles, while 2013 had no publications in the selected criteria.…”

“…In this existing (Franco et al, 2021) research, various deep‐learning (auto‐encoders) outcome measurements were performed for different cancer isoform detection. Validating cancer isoforms on 4 cancer types taken from TCGA (The Cancer Genome Atlas) datasets that use 4 auto‐encoder execution (Chen et al, 2018) (Huang et al, 2018) (Manica et al, 2019; Tan et al, 2019; Vougas et al, 2019; Xia et al, 2018; Zhang et al, 2018; Zhang et al, 2021; Zhong et al, 2018).…”