A novel analysis strategy for integrating methylation and expression data reveals core pathways for thyroid cancer aetiology

Table 1 Analysis models and their short descriptions.

Analysis Models	Model Descriptions (FDR<0.01 for all models)
Model 1	Only differentially expressed genes
Model 2	Only differentially methylated regions
Model 3	Differentially expressed and differentially methylated genes
Model 4	Significant genes when methylation and expression significances combined
Model 5	Genes with more than 15% methylation change and inversely correlated with expression
Model 6	Genes with more than 15% methylation change
Model 7	Significant genes with methylation level change more than 15% and obtained after combining methylation and expression significance values
Model 8	Significant genes with methylation level change more than 15%, inversely correlated with expression and obtained after combining methylation and expression significance values
Model 9	Significant genes with methylation level change more than 15%, positively correlated with expression and obtained after combining methylation and expression significance values
Model 10	Significant genes with methylation level change more than 40% and obtained after combining methylation and expression significance values

Short summary of each analysis model. For the search of finding the optimal analysis strategy, we have applied 10 different analysis models on different data selection options.

ISSN: 1471-2164