Table 2 Accuracy of Two-Transcript Classifiers on Diverse Phenotypes
From: Two-transcript gene expression classifiers in the diagnosis and prognosis of human diseases
Classification Task | Accuracy (Sens./Spec.) | Classifier Gene Pair and Annotated Functions | False Discovery |
|---|---|---|---|
GIST/LMS | 100% (100.0/100.0) | PRUNE2 (Regulation of Apoptosis) OBSCN (Muscle Differentiation & Signaling) | < 10 E-5 |
Crohn's Disease | 96.04% (96.6/95.2) | TBX21 (Immune Modulation) APOLD1 (Angiogenesis; Lipid Metabolism) | < 10 E-5 |
Cardiomyopathy | 74.23% (58.1/87.0) | PDE8B (Phosphodiesterase; cAMP Metabolism) ZNF263 (Zinc-Finger Transcription Factor) | < 0.002 |
Type I Diabetes | 91.43% (96.3/75.0) | CD1D (Antigen Processing and Presentation) PSD (ARF/RAS Signal Transduction) | < 0.002 |
Type II Diabetes | 100% (100.0/100.0) | UNC5A (Regulation of Apoptosis) ATG16L2 (Protein Transport; Autophagy) | < 0.005 |
UC Transformation | 96.3% (81.8/100.0) | PAK2 (Kinase Signaling; Cell Cycle Regulation) FLT3LG (Immune Activation) | 0.05910 |
Gram-Negative/Viral | 100% (100.0/100.0) | CD40 (Immune Response; B Cell Proliferation) SETD6 (Histone Methyltransferase Activity) | < 10 E-4 |
HIV Infection | 100% (100.0/100.0) | GAD1 (Glutamic Acid Metabolism) RHD (Erythrocyte Function) | < 10 E-4 |