Table 3.
Multi-omic EWAS studies identifying marker networks in PDAC.
| Source | Tissue Type | Techniques | Sample Size | Comparisons | Strengths | Weakness |
|---|---|---|---|---|---|---|
| [121] | Pancreas tissue | 5617 miRNA—Affymetrix GeneChip miRNA 3.0 | 104 PDAC and 17 benign pancreas tissue | Cancer vs. benign | ● Candidate markers annotated using gene ontology analysis | ● New approach - unvalidated |
| 33,297 mRNA—HuGene 1.0 ST | Validation in GEO and TCGA databases | Cancer vs. normal | ● Selection of genes based on predictive measures and adjusted p-values | ● Weights are dataset dependent, however, limited markers to validation in at least 2 datasets | ||
| [134] | PDAC tumor tissue and cell lines | exome—llumina HiSeq 2000) | 3 different cell lines and 6 primary pancreatic cancer tumors | Primary tumor vs cell lines | ● Combined exome data and transcriptome data | ● Variant analysis and interpretation |
| transcriptome—RNA-seq (Illumina HiSeq 2000) | ● Variant filtering in pipeline removes most false positives | ● Biopsy samples generally included normal tissue | ||||
| ● Made analysis pipeline available for others to try and establish standard and reproducibility | ● Exome only on cell lines | |||||
| [122] | Pancreas tissue | multiple—Table 1 in reference | Cancer vs. normal | ● Used FDR to determine significance | ● Datasets with no class-based clustering were excluded | |
| Survival | ● Focused meta-analysis on functional markers | ● Several arbitrary filters applied - currently no standardized data combining techniques | ||||
| ● Visualization of significant results | ● Clinical factors not taken into account in survival plots | |||||
| ● Large sample size - meta analysis | ● Hard to identify causal changes | |||||
| [135] | Cell lines | Agilent Human Whole-genome expression microarray | 3 BxPC-3 and 3 BxPC-3ER | Treatment response | ● Investigated specific expression changes associated with erlotinib resistance using BXPC cell line | ● Understanding metabolite changes is limited |
| ● Identified potential metabolic pathways and associated genes to target to counter resistance | ● Expression and phosphorylation of RTKs not consistent with previous reports |