Quality control checkpoints for high throughput DNA methylation measurement using the human MethylationEPICv1 array:...

The performance of FFPE tissue-derived DNA on the MethylationEpicV1.0 array can be unpredictable as the protocol only has two quality checks; checkpoint 1 for DNA quantity and checkpoint 2 for DNA quality assessment. We sought to incorporate a third, previously developed bisulfite conversion quality check prior to processing 255 FFPE tissue derived DNA samples.

FFPE tissue-derived DNAs were prepared for 255 prostate tumour specimens and four controls. Checkpoint 1 assessed all samples to have 500ng DNA available for analysis except for two samples that yielded 483 and 486ng. All DNA samples passed both the quality checkpoint 2 and the bisulfite conversion quality assessment checkpoint 3. Assessment of array performance showed one of the 259 (0.4%) DNA samples had less than 90% of probes detected at p = 0.05. Controls and replicate showed reliable reproducibility with correlations of 0.981 and 0.994. High quantity and quality measures of the FFPE tissue-derived DNAs (assessed by checkpoint 1 and 2) were likely responsible for 99.6% of DNAs producing high quality EPIC array data. This report suggests that checkpoint 3 has limited value in a setting where the FFPE tissue derived DNA has high quantity and quality measures.