Table 8 Data-driven precision diagnosis in digestive diseases based on integrated omics
From: Data-driven decision-making for precision diagnosis of digestive diseases
First author, year | Disease | n | Data source and specific task | ML method | Diagnostic performance | Refs. |
|---|---|---|---|---|---|---|
Liu, 2016 | HCC | 256 | CNAs, DNA methylation, mRNA, and miRNA data from TCGA; subtyping HCC by multi-omics data | PCA/LR/Clustering | AUC: 0.780–1.000 | [175] |
Al-Harazi, 2021 | CRC | 89 | Whole-genome gene expression profiling and CNA data sets from GEO database; classification based on the cores of 15 subnetwork markers | SVM/PCA/Clustering | Accuracy: 98.0% | [176] |
Hoshino, 2022 | CRC | 24 | Radiomics data of CT image and DNA sequencing data of tumor mutation burden; prediction of tumor mutation burden based on the image features | RF/XGBoost | Accuracy: 68.2% | [177] |
Gawel, 2019 | CRC | 160 | Public proteomics and transcriptomics data sets of tumor and adjacent tissues; classification based on nine secreted protein markers | Random elastic net | Sensitivity: 90.0% Specificity: 92.0% | [178] |
Gai, 2023 | CAG/GC | 319 | Fecal metabonomics and microbiota profiles data; classification based on 2 fecal metabolites and 2 gut microbes | SVM/RF | AUC: 0.88 Accuracy: 85.7% | [179] |
Huang, 2022 | CRC | 743 | Genomic and epigenetic profiles data sets of tissues from TCGA and GEO databases; classification based on DNA methylation and mutation burden data | LASSO/SVM/PCA/LR | AUC: 0.857–1.000 | [180] |
Cao, 2020 | CRC | 1214 | Pathomics, genomic and transcriptomic data sets; classification base on pathomics signature | Residual CNN/XGBoost/Naive Bayes | AUC: 0.850–0.885 | [181] |
Gonzalez, 2022 | Crohn disease | 182 | Fecal metaproteomics, metagenomics, metabolomics, and host genetics data; prediction of CD location based on a multi-omics feature set from metabolomics and metaproteomics | RF/LR/ExtraTrees/DT/Naive Bayes/KNN/SVC/MLPC/Voting Classifier/Adaboost | AUC: 0.94 | [182] |
Adel-Patient, 2023 | EoE | 32 | Tissue transcriptomics, tissue and blood immunologic components, and plasma metabolomics data sets; classification based on combining plasma metabolomics and cytokine biomarkers | PLS–DA/PCA | AUC: 0.929 | [183] |
Xing, 2023 | CRC | 212 | Tissue transcriptomics and plasma metabolomics data; classification based on combining metabolomics and RNA-seq data | PLS–DA/PCA | AUC: 0.904–0.923 | [184] |
Kel, 2019 | CRC | 202 | Full genome gene-expression data and genomic CpG island methylation data from tumor and gut epithelial tissues; classification based on 6 hypermethylated gene markers | F-Match/CMAcorrel/SVM/master-regulator search algorithm | Accuracy: 92.3% | [185] |
Ding, 2019 | CRC | 315 | Transcriptomics and proteomics data of CRC; classification based on secreted biomarkers | SVM | Accuracy: 85.9% | [186] |