Deep learning-based clustering approaches for bioinformatics

Abstract

Clustering is central to many data-driven bioinformatics research and serves a powerful computational method. In particular, clustering helps at analyzing unstructured and high-dimensional data in the form of sequences, expressions, texts and images. Further, clustering is used to gain insights into biological processes in the genomics level, e.g. clustering of gene expressions provides insights on the natural structure inherent in the data, understanding gene functions, cellular processes, subtypes of cells and understanding gene regulations. Subsequently, clustering approaches, including hierarchical, centroid-based, distribution-based, density-based and self-organizing maps, have long been studied and used in classical machine learning settings. In contrast, deep learning (DL)-based representation and feature learning for clustering have not been reviewed and employed extensively. Since the quality of clustering is not only dependent on the distribution of data points but also on the learned representation, deep neural networks can be effective means to transform mappings from a high-dimensional data space into a lower-dimensional feature space, leading to improved clustering results. In this paper, we review state-of-the-art DL-based approaches for cluster analysis that are based on representation learning, which we hope to be useful, particularly for bioinformatics research. Further, we explore in detail the training procedures of DL-based clustering algorithms, point out different clustering quality metrics and evaluate several DL-based approaches on three bioinformatics use cases, including bioimaging, cancer genomics and biomedical text mining. We believe this review and the evaluation results will provide valuable insights and serve a starting point for researchers wanting to apply DL-based unsupervised methods to solve emerging bioinformatics research problems.

Data mining application to healthcare fraud detection: a two-step unsupervised clustering method for outlier detection with administrative databases

“…There are several clustering algorithms that may be considered as alternatives to the k-means. Among others, several different hierarchical clustering methods (both agglomerative and divisive) [28,30], together with more recent efficient and effective approaches such as DBSCAN [31], spectral methods [32], or the modern Deep Learning-based approaches [33]. All these algorithms present peculiarities that allow them to separate at best the clusters intrinsically available in data, up to a different extent of complexity in the shape of the aforementioned clusters.…”

Topological Feature Extraction and Visualization of Whole Slide Images using Graph Neural Networks

“…Patch-level features may be extracted using pretrained CNNs such as ImageNet, which has learned a huge collection of convolutional filters and features correspondent to 1000 common objects such as dogs, cats and birds 25 . Features may also be acquired using unsupervised approaches such as variational autoencoders (VAE) 26 or self-supervised techniques such as contrastive predictive coding (CPC) 14 or SimCLR 27 . Finally, patch-level features may be learned after pretraining on histology targets of interest, such as classified objects or ROI.…”

BCD-WERT: a novel approach for breast cancer detection using whale optimization based efficient features and extremely randomized tree algorithm

“…Humpback whales are the largest group of baleen whales and they usually spend their days as a group. They hunt small groups of krill and small fishes close to the surface by creating bubbles along a spiral path around their prey and then they swim up to the surface following this path (Mirjalili & Lewis, 2016;Karim et al, 2020). Encircling Prey: To hunt, humpback whales can identify the location of their prey and attack them by encircling them.…”