Developing an Original Multi-Omics Biomarker Panel for Early Detection of Myelodysplastic Syndrome in Nigerian Adults: Integrating Proteomics, Metabolomics, and Transcriptomics

This dissertation develops an original multi-omics biomarker panel for early detection of myelodysplastic syndrome (MDS) in Nigerian adults by integrating plasma proteomics, serum metabolomics, and peripheral blood transcriptomics, addressing the critical unmet need for sensitive non-invasive MDS detection tools in populations where late-stage diagnosis remains the norm. MDS is substantially underdiagnosed in Nigeria due to dependence on bone marrow histology as the sole diagnostic route, delaying treatment and worsening outcomes. A discovery-validation design recruited 120 confirmed MDS cases (WHO 2022 classification), 80 cytopenic non-MDS controls with other explanations, and 80 healthy controls at Lagos University Teaching Hospital and University of Nigeria Teaching Hospital. Discovery phase used mass spectrometry-based plasma proteomics (1,800 proteins quantified), NMR-based serum metabolomics, and peripheral blood 3-prime RNA sequencing. Multi-omics data integration was performed by multi-block sparse partial least squares discriminant analysis. An original 18-feature biomarker panel incorporating three proteomic markers, four metabolites, and 11 transcripts achieved AUC of 0.89 for MDS detection versus non-MDS cytopenias in cross-validated discovery and 0.86 in an independent validation cohort. The panel outperformed cytomorphology alone (AUC 0.73). The dissertation proposes the Nigerian MDS Multi-Omics Detection Index as an original diagnostic contribution, with clinical utility demonstrated in differential diagnosis of unexplained cytopenia. Keywords: myelodysplastic syndrome, multi-omics, biomarker panel, proteomics, early detection.