A decomposition model to track gene expression signatures: preview on

A decomposition model to track gene expression signatures: preview on

observer-independent classification of ovarian cancer

Ann-Marie Martoglio^1*+, James W. Miskin²⁺,

Stephen K. Smith¹ and David J. C. MacKay*²

¹Reproductive Molecular Research Group, Department of Pathology and Department of Obstetrics and Gynaecology, University of Cambridge, Tennis Court Road, Cambridge, CB2 1QP, U.K..

²Cavendish Astrophysics Group, Cavendish Laboratory, University of Cambridge, Madingley Road,Cambridge, CB3 0HE, U.K..

+These authors contributed equally to this work.Ê

*Correspondence mayÊ be addressed to amm53@cam.ac.uk or mackay@mrao.cam.ac.uk

Links referenced in the above article:

Ovarian cancer data (17 tissues, 175 genes - as a tab-delimited text file for downloading - Click to View, Right-click to Download)

Learning run (independent component analysis, (ICA))

Fixed matrix tests (post-ICA learning run test)

Table 1 (signal-to-noise ratios for emerging gene signatures)

Table 2 (leading genes in emerging gene signatures)

For an article describing the assumptions underlying the latent-variable-modelling work of Miskin, Martoglio and MacKay for microarrays please visit the following link;

http://www.inference. phy.cam.ac.uk/mackay/abstracts/icagenes.html