Paul Skipp

Global gene expression data is a high resolution image of molecular phenotype. Although each molecule is measured with low accuracy, topological measurement of shape works like biology – the data points are noisy, high dimensional and with a lot of redundancy built in. Biological shapes are more complicated than the examples shown here (the star and the rectangle), and TopMD is happy to deal with these complex shapes. Here, we illustrate how three discrete biomarkers do not distinguish between the two shapes, and with variability in populations and measurement, the discrete biomarkers are even less successful in discriminating shapes. We show on the right hand side TopMD’s approach, where high dimensional noisy data accurately discriminates complex shapes. TopMD uses biological pathways as coordinates for measuring the shape of global gene expression, accurately characterizing molecular phenotypes. TopMD offers revolutionary accuracy in identifying disease phenotype, the shape of global gene expression. Discrete markers of disease are subject to failure as predictive biomarkers due to variability in study populations, patient populations and in their measurement precision and accuracy. Discrete markers of disease identified by differential gene expression analysis are subject to variability in measurement accuracy and variability across a population. These markers can be false positives, or their variability means they cannot be developed into successful molecular diagnostics. Disease is characterised by modulation of biological pathways. TopMD measures the shape of differentially modulated pathways, robustly diagnostic of disease phenotype.