Ensemble analyses improve signatures of tumour hypoxia and reveal inter-platform differences

• Published in: BMC bioinformatics Vol. 15; no. 1; p. 170
• Main Authors: Fox, Natalie S, Starmans, Maud H W, Haider, Syed, Lambin, Philippe, Boutros, Paul C
• Format: Journal Article
• Language: English
• Published: England BioMed Central Ltd 06.06.2014; BioMed Central
• Subjects: Algorithms; Biomarkers; Breast cancer; Cancer therapies; Cell Hypoxia; Classification; Colleges & universities; Gene expression; Gene Expression Regulation, Neoplastic; Genetic aspects; Genomes; Humans; Hypoxia; Medical prognosis; Medical research; Neoplasms - pathology; Oncology; Patients; Physiological aspects; Prognosis; Reproducibility of Results; Tissue Array Analysis; Tumors

The reproducibility of transcriptomic biomarkers across datasets remains poor, limiting clinical application. We and others have suggested that this is in-part caused by differential error-structure between datasets, and their incomplete removal by pre-processing algorithms. To test this hypothesis,...