Computational protocol: Global DNA methylation patterns in Barrett’s esophagus, dysplastic Barrett’s, and esophageal adenocarcinoma are associated with BMI, gender, and tobacco use

Similar protocols

Protocol publication

[…] Pathway enrichment analysis of significantly differentially methylated genes between any two sample groups was performed using pathway definitions derived from the NCI Pathway Interaction Database (NCI-PID), a curated collection of known biomolecular interactions and key signaling pathways associated with cancer []. The enrichment analysis was performed using the hypergeometric test to evaluate if genes belonging to a given pathway were enriched among the significantly differentially methylated loci. We elected to increase the possibility that altered molecular pathways would be biologically relevant by restricting our NCI-PID analysis to only cancer-associated DML. Hochberg FDR methodology and pathways with FDR ≤ 0.05 were considered significantly methylated genes. This was followed by followed by assessment of false discovery rate using the Benjamini Hochberg correction []. Genes with multiple differentially methylated probes were excluded if the methylation state of any probe was inconsistent between comparison groups. Gene set enrichment analysis was performed using genes from differentially methylated groups to identify affected Gene Ontology (GO) terms and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways across different comparison groups using hypergeometric testing provided by the GOstats software []. Probes whose target genes were not annotated to at least one GO term in the biological process ontology were filtered out. A gene set was considered altered if its number of differentially methylated CpG sites was higher or lower than expected using a p value <0.05. […]

Pipeline specifications

Software tools PID, GOstats
Databases KEGG
Application Protein interaction analysis
Organisms Nicotiana tabacum, Homo sapiens
Diseases Adenocarcinoma, Barrett Esophagus, Neoplasms