Cancers exhibit abnormal molecular signatures associated with disease initiation and progression. Molecular signatures could improve cancer screening, detection, drug development and selection of appropriate therapeutic interventions for individual patients. Typically only very small amounts of tissue are available from patients for analysis and biopsy samples exhibit broad heterogeneity that cannot be captured using a single marker. This presentation details application of in-house custom designed GenomeLab System multiplex gene expression assays. Details of design, tissue optimisation, total RNA quality and analytical and clinical validation will be discussed. Application to assess predictive gene signatures of normal, adenomatous polyp and carcinoma colon tissue using archived tissue bank material and an in-house custom designed human cell marker multiplex, the hCellMarkerPlex will be presented. The hCellMarkerPlex incorporates twenty-one gene markers: epithelial (EZR, KRT18, NOX1, SLC9A2), proliferation (PCNA, CCND1, MS4A12), differentiation (B4GANLT2, CDX1, CDX2), apoptotic (CASP3, NOX1, NTN1), fibroblast (FSP1, COL1A1), structural (ACTG2, CNN1, DES), gene transcription (HDAC1), stem cell (LGR5), endothelial (VWF) and mucin production (MUC2). Gene signatures distinguished normal, adenomatous polyp and carcinoma. Individual gene targets significantly contributing to molecular tissue types, classifier genes, were further characterised using real-time PCR, in-situ hybridisation and immunohistochemistry revealing aberrant epithelial expression of MS4A12, LGR5 CDX2, NOX1 and SLC9A2 prior to development of carcinoma. Identified gene signatures identify aberrant epithelial expression of genes prior to cancer development using in-house custom designed gene expression multiplex assays. This approach may be used to assist in objective classification of disease initiation, staging, progression and therapeutic responses using tissue biopsy material.