MetastaSite: Predicting metastasis to different sites using deep learning with gene expression data

Abstract

Deep learning has massive potential in predicting phenotype from different omics profiles. However, deep neural networks are viewed as black boxes, providing predictions without explanation. Therefore, the requirements for these models to become interpretable are increasing, especially in the medical field. Here we propose a computational framework that takes the gene expression profile of any primary cancer sample and predicts whether patients’ samples are primary (localized) or metastasized to the brain, bone, lung, or liver based on deep learning architecture. Specifically, we first constructed an AutoEncoder framework to learn the non-linear relationship between genes, and then DeepLIFT was applied to calculate genes’ importance scores. Next, to mine the top essential genes that can distinguish the primary and metastasized tumors, we iteratively added ten top-ranked genes based upon their importance score to train a DNN model. Then we trained a final multi-class DNN that uses the output from the previous part as an input and predicts whether samples are primary or metastasized to the brain, bone, lung, or liver. The prediction performances ranged from AUC of 0.93–0.82. We further designed the model’s workflow to provide a second functionality beyond metastasis site prediction, i.e., to identify the biological functions that the DL model uses to perform the prediction. To our knowledge, this is the first multi-class DNN model developed for the generic prediction of metastasis to various sites.

A review and analysis of key biomarkers in Alzheimer’s disease

“…Researchers are also dedicated to using artificial intelligence (AI), machine learning (ML), and deep learning (DL) algorithms for detecting AD and integrating different types of data (Alamro et al, 2023). These data types include, but are not limited to, neuroimaging data, non-coding RNAs, transcriptomic data (Qorri et al, 2020), miRNA biomarkers , or other genomic data (Monk et al, 2021).…”

Identification of hub genes and potential molecular mechanisms related to drug sensitivity in acute myeloid leukemia based on machine learning

“…Advancement in machine learning (ML) techniques and methods is fueling drug discovery and healthcare research in a large way. ML algorithms are extensively used in today's healthcare research for disease diagnosis, discovering potential prognostic biomarkers and drug targets in various pathophysiological conditions starting from viral infections to neurodegeneration disorders (Barman et al, 2019;Alamro et al, 2023;Taheri and Habibi, 2023;Turki and Taguchi, 2023). A few of the popular ML techniques/algorithms used in biological research include support vector machine (SVM) (Noble, 2006), artificial neural network (ANN), random forest (RF), and gradient boosting tree (GBT).…”

“…A few of the popular ML techniques/algorithms used in biological research include support vector machine (SVM) (Noble, 2006), artificial neural network (ANN), random forest (RF), and gradient boosting tree (GBT). Alamro et al, (Alamro et al, 2023), used ranking and feature selection methods to first shortlist the hub genes associated with Alzheimer's disease (AD) and then employed ML and deep learning (DL) methods to differentiate between AD patients and healthy controls using the selected gene-sets. Taheri et al, Taheri and Habibi, (2023) focused on a more recent problem and used three different unsupervised learning algorithms to rank the important genes and finally identified a set of 18 key genes related to COVID-19 disease.…”

OncoRTT: Predicting novel oncology-related therapeutic targets using BERT embeddings and omics features

“…Thus, several groups have proposed models developed with machine learning (ML) and deep learning (DL) techniques to address cancer-related issues. These models integrate features of the biological processes to accomplish various tasks, including identifying new gene-disease associations, pinpointing the cancer driver genes ( Althubaiti et al, 2019 ; Althubaiti et al, 2021 ), predicting cancer-specific biomarkers ( Pal et al, 2007 ; Tabl et al, 2019 ), predicting anticancer peptides ( Arif et al, 2022 ), and predicting pan-cancer metastasis ( Albaradei et al, 2019 ; Albaradei et al, 2021a ; Albaradei et al, 2021b ; Albaradei et al, 2022c ) ( Albaradei et al, 2022 ). There are also other models focused on cancer-related drug repurposing that predict drug response in cancer cell lines ( Liu et al, 2020 ) and novel oncology drug-target interactions (DTIs) ( Huang et al, 2016 ; Dezső and Ceccarelli, 2020 ).…”