Additional Supporting Information may be found in the online version of this article.

HEP_23488_sm_SupFig1.tif2677KSupporting Fig. S1. Schematic representation of the experimental protocol. (A) Male F344 rats were injected intraperitoneally with diethylnitrosamine (DENA, Sigma) at a dose of 150 mg/kg body weight. Following a 2-week recovery, rats were fed a diet containing 0.02% 2-acetylaminofluorene (2-AAF, Sigma) for 1 week followed by a two-thirds partial hepatectomy (PHx), and an additional week of 2-AAF diet. The animals were then returned to a basal diet, and euthanized at 10 weeks, 9, and 14 months in groups of three.
HEP_23488_sm_SupFig2.tif3058KSupporting Fig. S2. Morphology of the representative hepatic lesions in rats subjected to RH protocol. (A) Adenoma at 9 months. (B) Early HCC at 9 months. (C) HCC at 14 months after DENA initiation. Inset shows multiple-cell thick plates of hepatocytes with atypical nuclei. H&E staining. Original magnification: X100 (A-C), insert, X200.
HEP_23488_sm_SupFig3.tif10613KSupporting Fig. S3. Expression of GSTP and CK19 in focal lesions. Rats were treated with DENA and 2-AAF, and sacrificed 10 weeks after initiation. Parallel liver sections were immunostained for GSTP (A, C, E) and CK19 (B, D, F). GSTP+ nodules displayed different degrees of CK19 expression. Original magnification, X40.
HEP_23488_sm_SupFig4.tif7949KSupporting Fig. S4. Supervised analysis of functional gene connectivity. Ingenuity Pathway Analysis (IPA) tools were used to analyze genes unique to each cluster. Genes upregulated in the CK19+ focal lesions as compared to normal liver are shown in red and downregulated genes are shown in green. Dashed lines represent indirect connectivity. The top 5 networks are involved in cancer, metabolism, and immune response (P<10−22 to 10−35). The significance is estimated for each network in IPA by determining the number of the differentially expressed genes within the network and the strength of each connection.
HEP_23488_sm_SupTab1.doc50KSupporting Table S1. Clinicopathological data
HEP_23488_sm_SupTab2.doc37KSupporting Table S2. Gene set overlaps with the Molecular Signatures Database (MSigDB)
HEP_23488_sm_SupTab3.doc182KSupporting Table S3. List of the 110 classifier genes
HEP_23488_sm_SupTab4A.doc49KSupporting Table S4A. Performance of the 110 gene classifier
HEP_23488_sm_SupTab4B.doc25KSupporting Table S4B. Performance of classifier during cross-validation
HEP_23488_sm_SupDocument.doc30KSupporting Methods

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.