Table 6.

Contributions of recent studies with manifold representation learning.

Author and Citation	Contributions
Zhang and Kiryu [81]	Developed MODEC, an unsupervised clustering method using manifold optimization and deep learning for identifying cancer subtypes.
Zhang et al. [82]	Developed the deep Bayesian perturbation Cox network (DBP) to effectively predict survival outcomes in cancer patients dealing with high-dimensional datasets.
Gupta et al. [83]	Developed continuous representation of codon switches (CRCS), a deep learning-based method for generating numerical vector representations of mutations with applications in detecting cancer-related somatic mutations and predicting patient survival.
Kim et al. [84]	Used a novel deep learning-based method to predict survival in oral cancer by analyzing tumor-infiltrating lymphocyte profiles.
Li et al. [85]	Developed CRCNet, a deep learning model for predicting survival outcome and the benefit of adjuvant chemotherapy in stage II/III colorectal cancer (CRC) patients.
Li et al. [86]	Employed deep learning to identify genetic mechanisms underlying immunosuppression in the survival of oral squamous cell carcinoma (OSCC) patients.
Shirazi et al. [87]	Developed a deep convolutional neural network (DCNN) for segmentation of whole-slide pathology images in glioblastoma to identify novel tumour cell–perivascular niche interactions associated with poor survival.
Skead et al. [88]	Conducted a deep learning and population genetics study on age-related clonal hematopoiesis (ARCH), demonstrating high accuracy in discriminating between evolutionary classes and captured signatures of purifying selection.
Wang et al. [89]	Utilized bidirectional long short-term memory (BiLSTM) to infer pan-cancer associated genes by examining the microbial model organism Saccharomyces Cerevisiae (Yeast) by homology matching.
Yin et al. [90]	Developed a convolutional neural network (CNN) model, named the CNN-Cox model, for survival prediction based on prognosis-related cascaded Wx feature selection.
Li et al. [91]	Constructed an immunophenotype-associated mRNA signature (IMriskScore) for predicting overall survival in patients with lower-grade glioma using deep learning neural networks with MRI radiomics.