Using Supervised Machine Learning for Prediction of Sjogren Syndrome in Salivary Glands Based on Gene Expression

Bulk RNA sequencing in minor salivary glands of Brazilian patients with primary Sjögren’s syndrome (SS) and healthy volunteers.
Available in GEO: GSE154926 (Contributor(s) Chiorini JA, Mo Y, Pranzatelli TJ, Michael DG, Ji Y, Rocha EM)

Author: Agustin Alejandro Martinez Chibly

Data Source for this project: https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE154926

Project Description: We will use supervised machine learning approach to predict sjogren's syndrome (ss) based on the transcriptional profile of salivary glands. We will test the performance of 6 different algorithms including logistic regression (LR), Linear Discriminant Analysis (LDA), K-Nearest Neighbors (KNN), Desicion Tree Classifier (DTC), Gaussian Naive Bayes (NB), and Support Vector Machines (SVM), using a k-fold cross-validation approach. The algorithms with the better performance will be tested in the validation set to predict SS patients.