Differential Gene-Expression Analysis with ML Validation

Which genes change in psoriatic skin, and does a classifier really confirm them? This project finds the differentially expressed genes between lesional psoriasis and normal skin in RNA-seq read counts (GEO GSE54456) with DESeq2, the Welch t-test and the Wilcoxon test, replicates them in an independent cohort (GSE121212) and validates them with random forest and logistic regression under nested cross-validation, permuted labels and random gene sets. 93 % of the DESeq2 genes replicate in another lab's patients; a perfect cross-validated AUC turns out to prove little (random genes reach 0.998), leaky gene selection scores 0.66 on pure noise, and the deployed 5-gene signature also flags atopic dermatitis.

Category: Bioinformatics - Transcriptomics and Gene-Expression Analysis

Transcriptomics & Gene-Expression Analysis Project B18 from ProTechSC Academy, with graphical abstract, video demonstration and project guidance.