Restore heavily degraded FFPE DNA

Find statistically relevant biomarkers in tissue that has been stored in paraffin for decades.

Measuring epigenetic markers in paraffin embedded (FFPE) blocks is challenging. Fixation extensively damages and shreds DNA. The nucleotides of these DNA fragments also become altered by processes such as depurination and de-amination.

The EpiRestore™ service restores heavily degraded DNA so it can be analyzed using Illumina’s MethylationEPIC micro-array. 
Moreover, our modified hybridization protocol allows you to start with as little as 200 ng of DNA. You can harvest a wealth of data of previously unattainable patient cohorts.

The advantages of EpiRestore™

  • Applicable for sample cohorts with low DNA integrity, such as paraffin blocks or Guthry cards
  • Restores DNA that is heavily degraded down to average lengths of 300 basepairs
  • Applicable for low amounts (200 ng) of genomic DNA
  • Gives biologically meaningful data: the data-set highly resembles the sample as if it was fresh frozen and stored without fixative
  • Excellent tool to screen large sample numbers due to cost-effective pricing and easy data-analysis.

DNA repair service

Our scientists have developed a proprietary DNA repair service that allows you to measure methylated and/or hydroxymethylated residues. This advanced method interrogates over 850,000 loci throughout the genome, selected by epigenetic experts. It delivers robust and highly reproducible results despite the very poor condition of the starting material.

The MethylationEPIC BeadChip contains >90% of the original 450K BeadChip content, allowing you to add FFPE sample data to previously analyzed cohorts.

Quality checks

We perform quality checks throughout the whole procedure, to ensure that only good quality results will be reported. We deliver all micro-array quality specs already loaded into the intuitive open-source software package MethylAID (van Iterson, M. et al. Bioinformatics 2014). You can immediately assess the quality of the data by scrolling through the graphs. Each describe a quality aspect: sample-dependent as well as sample-independent controls such as call-rates, hybridization controls and bisulfite conversion controls.

Guarantee for highly degraded FFPE samples

Our R&D scientists have processed many FFPE cohorts and have shown that we can extract biological meaningful data with great statistical certainty, even from highly degraded cohorts.

You retrieve the methylation status of 765,000 or more loci, from CpG sites selected by Epigenetics experts: a pan-enhancer and coding region view of the methylome, ready for use in epigenome-wide association studies on various human tissues.

What extent of DNA degradation can EpiRestore™ handle?

Four samples were selected that are representative for the range of samples that researchers encounter after DNA isolation from FFPE blocks. Additional types of DNA damage of the nucleotides -such as de-purination and oxidation- cannot be visualized by gel-electrophoresis.

Are data from restored samples similar to those of fresh frozen (FF) samples?

  • Tumors from a cohort dating from 1988, split in half and either fresh frozen or preserved as FFPE blocks
  • Average fragment size between 400 and 700 basepairs
  • EpiRestore™ service restored the FFPE DNA
  • Performed the fine-tuned protocol of the Illumina’s MethylationEPIC array to accommodate samples with low input material.

Each probe or CpG locus is represented multiple times on the array. The average of the fluorescent signals of each of these 850,000 loci (average beta-value or Avβ of each probe) is visualized in the figure as a black dot.


  • The highest correlation is observed using technical replicates of the fresh frozen sample
  • No EpiRestore™: no meaningful data can be extracted
  • With EpiRestore™: the two parts of one tumor that have been stored separately for almost 30 years show a concordance (R2) of 0.97, a resemblance usually observed in samples that are stored under excellent conditions.

What is the variation in the concordance between fresh frozen and FFPE?


  • EpiRestore™ greatly increases the concordance of severely degraded FFPE samples with their cognate Fresh Frozen aliquot
  • The more heavily damaged DNA, the greater the improvement of the results
  • EpiRestore™ significantly improves the biological relevance of the data in all samples

Prediction of the success rate


  • Cohort of samples, extracted from FFPE
  • Input concentration of all samples standardized on 200 ng or less
  • Average fragment size below 400 bp (Undamaged DNA: >20,000 bp)
  • EpiRestore™ service: DNA restoration followed by the adjusted MethylationEPIC protocol
  • qPCR analysis to quantitatively and qualitatively measure the DNA before starting the micro-array experiment.


  • Positive control consisting of good quality DNA: ~100% of the loci detected
  • The higher the Ct value on the x-axis, the lower the quality of the DNA
  • Our qPCR method shows a tight correlation between quality and call rates, the detected number of CpGs.
  • Even the most heavily degraded samples (fragment size peak at ~200 bp) show a call rate of 70%: the dataset of this sample contains the methylation status of >765,000 methylation sites (p>0.01).

The DNA restoration success rate can be measured by qPCR. This QC method can accurately predict the call rate: the amount of detected CpG loci on the micro-array that will be achieved.

With EpiRestore™ we generally expect samples to reach call rates of at least 90%. This ensures that you can detect and compare at least 765,000 loci that are selected by Epigenetics experts.

EpiRestore™ guarantee

For freshly isolated DNA, we routinely detect call rates of above 99%. We can guarantee to detect 95% of the loci with a p-value of p≤0.01). With our fine-tuned protocols we can start this procedure with 100 ng input or even less, in contrast to the 250 ng input recommended by Illumina.  Contact technical support to inquire.
The results obtained by EpiRestore™ are depending on the quality of the DNA. We guarantee that at least 90% of the CpGs are detected at p<0.01 for a minimum of 80% of the samples. Thus, at least ~765,000 CpG loci that show differential methylation patterns in e.g. health and disease, are called with a high level of statistical certainty.
This guarantee is valid for input amounts of at least 200 ng, enabling you to select only the regions of interest from the paraffin blocks.

Application Restoration


Percentage of detected CpGs guaranteed Min. no. of detected CpGs

(p ≤ 0.01)

For number of samples
Methylation No 95 % 807,500 100 %
FFPE Methylation Yes 90 % 765,000 80 %
Hydroxymethylation No 95 % 807,500 100 %
FFPE Hydroxymethylation Yes 90 % 765,000 80 %

Our scientists can give these guarantees because we have challenged our partners to send in the most degraded samples available. The most interesting cohort was especially selected for this purpose by a highly-regarded team of the Karolinska Institute. Even with fragment lengths of less than 300 bp (instead of the average DNA fragment size of 20,000 bp), the EpiRestore™ procedure delivered biologically relevant data, thereby fulfilling our GenomeScan guarantee.

Finding Biomarkers with consortium partners

Biomarkers are characteristics by which a disease can be identified or monitored. They reflect past environmental exposures, predict disease onset or course, or determine a patient’s response to therapy. GenomeScan partners now with many (academic) institutes to find diagnostic and predictive biomarkers for cancer and other diseases, based on epigenetic markers on the DNA. We participate in multiple consortia. Visit our research collaborations page to learn about how to successfully apply for funding together with us.

Reduced Representation & Whole Genome Bisulfite Sequencing


Targeted methylation

MethylationEPIC ‘850K’ BeadChip

EpiRestore: restore all low integrity DNA