 |
|
Roche NimbleGen provides the most sensitive and specific DNA methylation microarrays and services on the market. Affinity-based methods for enrichment of methylated DNA fragments, such as MeDIP (Methylated DNA immunoprecipitation) or the MBD (Methyl Binding Domain Protein) methods, are the preferred sample preparation methods for identifying methylated DNA on NimbleGen arrays. Coupled with one of these enrichment methods, our arrays and services offer:
Advantages 
Comprehensive Set of Array Designs Cover All Your Needs
Roche NimbleGen offers whole genome, promoter, and custom array designs, allowing you to choose the design that meets your research needs. Our whole-genome DNA methylation designs interrogate the non-repetitive regions of human, mouse, rat, and plant genomes at intervals ranging from <100bp to 255bp for unbiased discovery of methylated DNA regions. Arrays focused on biologically significant genomic regions are also available including promoter/CpG island array designs and promoter only designs aimed at well characterized promoter regions. For custom designs, researchers can specify their regions of interest for fine-tiling of genomic regions or promoters or design a tailored array with their own probe designs.
High-Resolution, High-Throughput Analysis of DNA Methylation
NimbleGen high density DNA methylation microarrays are available in three formats: 2.1M (2.1 million probes on a single glass slide), 3x720K (3 identical arrays per glass slide with 720,000 probes per array), 385K (385,000 probes on a single glass slide) and 4x72K (4 identical arrays per glass slide with 72,000 probes per array). The 2.1M high density microarrays enable coverage of the entire non-repetitive human genome at 100bp spacing using only 10 slides - greatly simplifying whole genome analysis. Now with the 3x720K and 4x72K arrays, you can increase sample throughput and lower the cost per sample for targeted DNA methylation analysis.
Figure 1
Figure 1: Detection of Differential DNA Methylation using the Three New Human DNA Methylation 2.1M Designs - MeDIP samples from OT (maternally-derived ovarian teratoma cells) and CHM (paternally-derived hydatidiform mole) cells were labeled with Cy5 and the control (input) DNAs with Cy3, pooled and technical replicates were hybridized to the three different human array designs. Raw data were analyzed using NimbleScan software and the P-values estimating the significance of enrichment were visualized using SignalMap software. The black horizontal bar indicate the regions tiled on the different array designs. The red box highlights the differentially methylated region between the OT and CHM samples, which was easily identified using all three array designs. The blue box highlights a region of DNA methylation detected on the whole-genome tiling design with 100bp probe spacing but not in the whole-genome economy design with 205bp probe spacing illustrating the greater detection sensitivity possible using arrays with 100bp probe spacing.
High Sensitivity and Specificity Provide Unparalleled Results
Roche NimbleGen's proprietary, light-mediated synthesis process produces high-density microarrays of long oligonucleotide probes (50-75mer). These long oligo arrays, when used in combination with high-stringency hybridization protocols, produce results of unparalleled sensitivity and specificity. In addition, because Roche NimbleGen performs DNA methylation array experiments using a two-color protocol, where input control and MeDIP samples are co-hybridized to the same array, inter-array variation is eliminated. As a result, NimbleGen DNA Methylation Arrays can readily detect as low as two methylated CpGs in a 500 bp fragment.
Easy-to-use Graphical View of Data Facilitates Discovery
Roche NimbleGen’s SignalMap software enables you to visually interpret your data and perform peak finding. SignalMap’s graphic representation of your data aids in the discovery of methylated or unmethylated DNA regions and well as differentially methylated DNA regions between samples. A free, 30-day demo version of SignalMap software is available for download.
Figure 2
Figure 2: Identification of DNA Methylation Using NimbleGen Arrays - From the raw data, NimbleGen generates scaled log2-ratio data for IP/input. A statistical method is used to generate p-value enrichment data for each probe; peaks (methylated regions) are then generated based on the p-value data. NimbleGen also provides gene annotations and CpG island annotations.
Most Up-to-date Genome Builds Ensure the Most Accurate Results
NimbleGen DNA Methylation designs are based on the latest genome assemblies and sequence annotations to ensure comprehensive and accurate representation of the genome. In addition, you can continue to access array designs based on past genome data builds, which can be particularly useful for comparisons to prior studies.
Array Designs
Roche NimbleGen offers a wide range of 2.1M, 3x720K, 385K and 4x72K DNA methylation array designs to choose from, or you can customize the array probe set to your specifications. The existing designs include whole genome survey sets, consisting of uniform tiling arrays covering all unique regions for multiple genomes, and a variety of promoter-based designs including annotated CpG islands plus promoters and promoter only designs. For custom designs, researchers can specify their regions of interest for a fine-tiling or promoter specific design or design their own targeted probes for a tailored array solution.
Whole-Genome Designs
- 2.1M Whole-Genome Sets tile uniformly across all unique, repeat-masked regions of the human, mouse and rat genomes. The 2.1M whole-genome sets offer a higher density format so that fewer arrays are needed to analyze these complex genomes at high resolution. Whole-genome sets are available in two formats: a 10-array set at 100bp probe intervals or the Economy 4-array sets at 200-255bp probe intervals. Each array in the human and mouse sets also include new positive, negative, and non-CpG control regions to aid in the assessment of array and sample performance.
- 385K Whole-Genome Sets tile uniformly across all unique regions of a genome at an average probe spacing of 100bp or less. A. Thaliana arrays are currently available.
Promoter Designs
- 2.1M Deluxe Promoter Arrays are single array designs for human and mouse that cover ~10 kb of all annotated promoters. In addition, these arrays tile through all annotated CpG islands, all annotated miRNA promoters, and manually selected ENCODE regions (human only). All regions are tiled at 100bp probe spacing. These arrays also include new positive, negative and non-CpG control regions to facilitate assessment of array and sample performance.
- 385K Two-Array Promoter Sets contain all annotated splice variants and alternative transcription start sites. This array set provides the most comprehensive tool for genome-wide mapping of transcriptional regulatory elements including all RefSeq genes, the Mammalian Gene Collection, and UCSC known genes to encompass the most comprehensive list of transcripts available. Designs are available for human, mouse, rat, and Arabidopsis thaliana.
- 385K RefSeq Promoter Arrays are single array designs containing all known well-characterized RefSeq genes (NM prefix). The promoter regions on these arrays are covered by 50-75mer probes with approximately 100bp spacing, dependent on the sequence composition of the region. Designs are available for human, mouse, and rat.
- 385K RefSeq XM Promoter Arrays are single array designs containing model reference sequences produced by NCBI's Genome Annotation Project (those accessions beginning with XM) that are predicted by genome sequence analysis. The entries represent ab initio predictions, or have some level of transcript or homology to known genes to support the gene predictions. They represent the transcripts and proteins that are annotated on the NCBI Contigs, and they may be different from GenBank submissions for mRNAs and/or the curated RefSeq records with NM, NR, NP accession prefixes. Designs are available for human and mouse.
- 385K Minimal Promoter Arrays are currently available for Arabidopsis thaliana.
Targeted Designs
- 3x720K CpG Island Plus RefSeq Promoter Arrays are multiplex slides with 3 identical arrays per slide enabling the simultaneous analysis of 3 DNA Methylation samples for increased sample throughput. These arrays cover the annotated CpG islands as well as the promoters of the well characterized RefSeq genes derived from the UCSC RefFlat files and are available for human, mouse and rat. These arrays also include new positive, negative and non-CpG control regions to help facilitate assessment of array and sample performance.
- 385K CpG Island Plus Promoter Arrays are single array designs that include all UCSC-annotated CpG islands and with the following promoter regions for all RefSeq gene promoter regions: 1kb (human) and 1.8kb (mouse and rat). DNA methylation positive control regions, such as the HoxA gene cluster, H19/IGF2 cluster, KCNQ1 cluster, and the IGF2R gene are also included on the arrays.
- 385K Targeted Tiling Arrays offer fine tiling (5 bp median probe spacing) of select biologically significant areas. The Human HOX Array tiles through the HoxA, HoxB, HoxC, HoxD, β-globin and Xist genes. The Human and Dog ENCODE arrays tile through the biologically significant ENCODE pilot regions (HoxA, β-globin, Apo among others). The Human ncRNA (non-coding RNA) Array tiles through the promoters of the 462 miRNA genes, 362 snoRNAs and 89piRNA clusters as well as selected ENCODE regions.
- 4x72K Human ENCODE Array offers a higher throughput, lower cost solution for testing the quality of DNA methylation samples before moving to large-scale studies, or for screening focused targeted regions of interest, for up to 4 independent samples per slide.
Custom Tiling Arrays
For a completely tailored DNA methylation array, customer specified genomic regions of interest are tiled with the same stringent probe selection methodology as used in NimbleGen’s catalog designs at the desired tiling density. Please contact Roche NimbleGen Sales for a quotation or contact Roche NimbleGen Technical Support with any questions about DNA methylation custom tiling arrays.
Array Formats
Customer Workflow Choices
DNA Methylation Delivery Workflow
Customers can purchase catalog arrays or custom arrays from NimbleGen, and perform the array experiments at their own laboratories or core facilities. NimbleGen arrays are synthesized on standard-sized glass microscope slides and are compatible with a range of hybridization, washing and scanning instrumentation. NimbleGen provides a complete user’s guide to support customers with sample processing, array hybridization, scanning, data extraction, and analysis. Please contact NimbleGen for a list of required equipment and reagents. Roche NimbleGen NimbleScan and SignalMap software enable the same data analysis and visualization to be performed as in the Array Service mode. A free, 30-day demo version of SignalMap software is available for download. NimbleGen also offers a training program to get you up and running with NimbleGen arrays quickly.
DNA Methylation Service Workflow
DNA methylation microarray service consists of the following steps:
- The customer selects a catalog DNA methylation array design or works with NimbleGen's Bioinformatics team to create a custom array design. NimbleGen manufactures the array.
- The customer prepares their samples, including immunoprecipitation with anti-5-methyl-cytidine (MeDIP, recommended) or the MBD protein method. The DNA is amplified (if required) according to recommended protocols and samples are shipped to NimbleGen's Service Lab.
- NimbleGen labels the samples, performs the hybridization, scans the array, extracts the data, and performs a preliminary data analysis.
- The customer receives the raw data, scaled log2-ratio data, p-value data, peaks data (methylated regions), genome annotation, reports listing peaks with associated transcripts and gene annotation information, and complete NimbleGen array design documentation.
Sample Requirements for 2.1M Arrays
| Sample Required |
6.5μg each IP and control samples |
| Sample Concentration |
250-500ng/μl |
Sample Requirements for 3x720K and 385K Arrays
| Sample Required |
4μg each IP and control samples |
| Sample Concentration |
250-500ng/μl |
Availability
Delivery and Service Workflows are not available for every DNA Methylation array design. Please consult the Availability Guide on this page for a complete list.
Availability Guide
References
| Applications |
References |
| DNA Methylation Defines Repression of Pluripotency in Progenitor Cells |
| Using a murine system that progresses from stem cells to lineage-committed progenitors to terminally differentiated neurons, DNA methylation and Polycomb-mediated histone H3 methylation (H3K27me3) were analyzed. Several hundred promoters, including pluripotency and germline-specific genes, become DNA methylated in lineage-committed progenitor cells, suggesting that DNA methylation may already repress pluripotency in progenitor cells. Loss and acquisition of H3K27me3 was detected at additional targets in both progenitor and terminal states. Many neuron-specific genes that become activated upon terminal differentiation are Polycomb targets only in progenitor cells. Promoters marked by H3K27me3 in stem cells frequently become DNA methylated during differentiation. These data suggest a model how de novoDNA methylation and dynamic
switches in Polycomb targets restrict pluripotency and define the developmental potential of progenitor cells. |
Mohn F, et al. “Lineage-Specific Polycomb Targets and De Novo DNA Methylation Define Restriction and Potential of Neuronal Progenitors,” Mol Cell. 2008 Jun 20;30(6):755-66. [Epub 2008 May 29]. |
| Promoter DNA Methylation in the Human Genome |
| DNA methylation, RNA polymerase occupancy, and histone modifications were measured at 16,000 promoters in primary human somatic and germline cells. CpG-poor promoters are found to be hypermethylated in somatic cells, which do not preclude their activity. This methylation is present in male gametes and resulted in evolutionary loss of CpG dinucleotides. In contrast, strong CpG-island promoters are mostly unmethylated, even when inactive. Weak CpG-island promoters are distinct, as they are preferential targets for de novomethylation in somatic cells. These results show that promoter sequence and gene function are major predictors of promoter methylation states. Moreover, it was observed that inactive unmethylated CpG-island promoters show elevated levels of dimethylation of Lys4 of histone H3, suggesting that this chromatin mark may protect DNA from methylation. |
Weber M, et al. “Distribution, silencing potential and evolutionary impact of promoter DNA methylation in the human genome,” Nature Genetics2007; 39(4):457-466. |
| Interdependence between Methylation and Transcription in Arabidopsis thaliana |
| DNA methylation was mapped in the entire Arabidopsis thalianagenome at high resolution. DNA methylation covers transposons and is present within a large fraction of A. thalianagenes. Methylation within genes is conspicuously biased away from gene ends, suggesting a dependence on RNA polymerase transit. Genic methylation is strongly influenced by transcription: moderately transcribed genes are most likely to be methylated, whereas genes at either extreme are least likely. In turn, transcription is influenced by methylation: short methylated genes are poorly expressed, and loss of methylation in the body of a gene leads to enhanced transcription. These results indicate that genic transcription and DNA methylation are closely interwoven processes. |
Zilberman D, et al. “Genome-wide analysis of Arabidopsis thaliana DNA methylation uncovers an interdependence between methylation and transcription,” Nature Genetics2007; 39(1):61-69. |
| Genome-wide DNA Methylation of CpG Islands |
| Tiling arrays were used in combination with the methylated CpG island recovery assay (MIRA) to investigate methylation of CpG islands genome-wide. All four HOX gene clusters were found on chromosomes 2, 7, 12, and 17, which are preferential targets for DNA methylation in cancer cell lines and in early-stage lung cancer. Altogether, more than half of all CpG island-associated homeobox genes in the lung cancer cell line A549 were methylated. Comparison with ENCODE-derived data shows that the lack of methylation at CpG- rich sequences correlates with presence of the active chromatin mark histone H3 lysine-4 methylation in the HOXA region. Methylation analysis of HOXA genes in primary squamous cell carcinomas of the lung led to the identification of the HOXA7- and HOXA9-associated CpG islands as frequent methylation targets in stage 1 tumors. Widespread methylation of homeobox genes lends support to the hypothesis that a substantial fraction of genes methylated in human cancer are targets of the Polycomb complex. |
Rauch T, et al. “Homeobox gene methylation in lung cancer studied by genome-wide analysis with a microarray-based methylated CpG island recovery assay,” Proc. Natl. Acad. Sci. USA2007; 104(13):5527- 5532. |
| Large-scale Mapping of MeCP2-binding Sites |
| Mutations in MECP2 cause the autism-spectrum disorder Rett syndrome. MeCP2 is predicted to bind to methylated promoters and silence transcription. However, the first large-scale mapping of neuronal MeCP2-binding sites on 26.3 Mb of imprinted and nonimprinted loci revealed that 59% of MeCP2-binding sites are outside of genes and that only 6% are in CpG islands. Integrated genome-wide promoter analysis of MeCP2 binding, CpG methylation, and gene expression revealed that 63% of MeCP2-bound promoters are actively expressed and that only 6% are highly methylated. These results indicate that the primary function of MeCP2 is not the silencing of methylated promoters. |
Yasui DH, et al. “Integrated epigenomic analyses of neuronal MeCP2 reveal a role for long-range interaction with active genes,” PNAS, 2007; 104:19416-19421 |
Software
NimbleGen NimbleScan software for data extraction provides tools specifically developed for detecting regions of DNA methylation from array data generated using MeDIP or other affinity-based method (such as MBD). For each array, these tools generate a list of methylated regions (peaks) and produce reports that map peaks to specific gene promoters. A free, 30-day demo version of NimbleScan software is available for download.
NimbleGen SignalMap software enables you to visually interpret the scaled log2-ratio, p-value, and peak data generated by NimbleScan. Graphical representation of your data aids the discovery of epigenetic and transcriptional regulation, including DNA hypermethylation within a target genome and the correlation of promoter and genic methylation with gene expression. A free, 30-day demo version of NimbleScan software is available for download.
Roche NimbleGen Literature
For a complete listing of literature covering all Roche NimbleGen products and services please visit our literature page.
FAQ
| Hide All Topics Show All Topics |
| Sample Preparation |
| What kind of samples can be used on NimbleGen's DNA methylation arrays? |
These arrays are designed to detect differences between a sample enriched for methylated DNA and a control sample, such as total genomic DNA. We recommend samples enriched using affinity-based methods that utilize an anti-5-mC antibody or MBD proteins. We also accept enzyme based methods to detect DNA methylation, such as Hpa II tiny fragment enrichment by ligation-mediated PCR (HELP). |
| Do you have a recommended protocol for front-end sample processing for producing DNA fragments enriched for methylation? |
Yes, we recommend the MeDIP (methylated DNA immunoprecipitation) protocol (Nat Genet. 2005 Aug;37(8):853-62) followed by amplification if necessary. Please contact NimbleGen's Technical Services for a detailed sample preparation protocol. |
| What is the minimum amount of DNA required to perform a NimbleGen recommended MeDIP experiment? |
We recommend starting with 5μg high-quality genomic DNA. However, even smaller amount DNA can be used (1μg or even 200ng). The IP process generally yields 5-10% of the original starting DNA, and you can amplify your IP-ed DNA using whole genome amplification (WGA; kit available from Sigma) in order to obtain at least 4μg of DNA. |
| Do I need to amplify my MeDIP samples? |
We are able to use unamplified samples for labeling reactions if at least 2.5μg of enriched methylated DNA was obtained. Greater amounts of enriched DNA can be obtained by starting with more DNA and by pooling samples from multiple experiments. If amplifcation is necessary, we recommend using the WGA kit (Sigma, catalog #WGA2-50RXN) for MeDIP samples. We have routinely observed that the WGA method introduces little bias during the amplification. |
| What is the effect of CpG content on the MeDIP reaction? |
Studies by Weber et al. (Nat. Genet. 2005, 37(8):853-862) have shown that the more more methylated CpG dinucleotides within a given fragment, the more DNA that is immunoprecipitated. The increase immunoprecipitated material is due to the 5-methylcytidine antibody having more available epitopes (methylated CpG). |
| What are the sample requirements for DNA methylation arrays? |
We require 4μg DNA (although 2.5μg is acceptable) at a concentration of 250-500ng/μl with the majority of fragments greater than 200bp. The A260/A280 ratio should be at least 1.7 and the A260/A230 ratio should be at least 1.6. |
| Does NimbleGen perform front-end sample processing (e.g. IP and amplification)? |
No, NimbleGen is not currently set up to perform contract DNA methylation experiments for our customers. |
| Should I use a negative control and if so what should I use? |
Many of our customers do not use a negative control (e.g. nonspecific IgG antibody), but it is recommended if this is your first experiment with NimbleGen. You will want to have your negative control (IgG) co-hybridized with total DNA (input) in order to avoid a high signal/noise ratio that is often seen when the IgG sample is co-hybridized with the immunoprecipitated sample. |
| What should I use as a reference sample? |
The majority of our customers use total (input) sample as a reference. Using a nonspecific IgG is not a suitable reference. |
|
| Array Design |
| Do we have a catalog CpG island array? |
Yes, we now offer our Human and Mouse 2.1M Deluxe Promoter Arrays for that include all annotated CpG islands and extensive coverage of all annotated gene and miRNA promoters. These arrays also include new positive, negative and non-CpG control regions to aid in assessment of experimental performance. (continue with previous info starting with) We also offer human, mouse, and rat CpG Island Plus Promoter arrays that include all UCSC annotated CpG islands plus 1kb human promoter or 1.8kb mouse promoter region of every RefSeq gene. Control tiled regions on the array include the HoxA cluster (methylation hotspot) and several imprinted loci, including H19/IGF2, KCNQ1, and IGF2R. |
| Are the probes designed from both strands? |
No, NimbleGen only designs probes based off of the forward strand. |
| How does NimbleGen address repetitive elements in the genome for DNA methylation designs? |
We have developed our own method of repeat masking which is dependent on the mean frequency of the 15-mers which make up each oligo. A table is made of the count of all 15-mers that appear in the genome, from both strands. Then a 15-mer window is slid along each oligo, looking up the count of each 15-mer in the table, and calculating the average count. A threshold is set, usually 100 for large eukaryotic genomes, and any probe that exceeds that threshold is eliminated from further consideration. Depending on the region of the genome being evaluated, approximately 20-25% of the DNA is excluded. For some designs we use conventional repeat masking, as done by the RepeatMasker program http://www.repeatmasker.org/). However, NimbleGen has no access to the repeat libraries necessary to use this application, so we rely on third parties to supply this type of masked sequence. We find, however, that RepeatMasker is often overly aggressive and can mask 50-55% of human DNA sequence. See the following paper for reference: Bioinformatics. 2006 Jan 15;22(2):134-41. WindowMasker: window-based masker for sequenced genomes; Morgulis A, Gertz EM, Schaffer AA, Agarwala R; National Center for Biotechnology Information, National Institutes of Health, Department of Health and Human Services Building 38A, Room 1003N, 8600 Rockville Pike, Bethesda, MD 20894, USA. |
| What spacing do you recommend for DNA methylation designs? |
We recommend probe spacing of 100bp or less. |
| Can I get a graphical representation of all probes for a cataloged design that I am interested in so I can see what regions of the genome have coverage? |
Yes, we can generate this information in GFF format for all of catalog designs. You will need a copy of Roche NimbleGen's SignalMap software to view the GFF files. A free, 30-day demo version of SignalMap is available for download. |
|
| Data Analysis |
| Does NimbleGen normalize DNA methylation data? |
No, there is no normalization of DNA methylation data. However, we do scale the GFF files by subtracting the bi-weight mean for the log-ratio values from each log-ratio value. This effectively centers the log-ratio values around zero. |
| How does NimbleGen perform analysis on DNA methylation data? |
After scaled log2 ratio data is generated a modified ACME algorithm (Method Enzymol. 2006; 411:270-282) is employed where a fixed-length window is slid along the length of each chromosome, testing at each probe using a one-sided Kolmogorov-Smirnov (KS) test whether the surrounding window is enriched for high-intensity probes relative to the rest of the array. Each probe has a corresponding p-value score (-log10) and a threshold is set to select regions that are enriched (i.e. methylated) in the test sample. If you are a DNA Methylation array customer and are stil using NimbleScan 2.3 you will need to download the ACME plug-in by clicking here. |
| Why is the data analysis for DNA methylation different from ChIP-chip? |
Since DNA methylation array data is often characterized by broad ranges of enrichment (whereas ChIP-chip data enrichment is more discrete) a different method to identify peaks from scaled log2-ratio data must be employed. |
| How do I validate a genome-wide DNA methylation array experiment? |
Common methods include bisulfite sequencing (single nucleotide resolution) to validate methylation status of all CpG dinucleotides within a peak called on the array, combined bisulfite restriction analysis (COBRA, determine percent methylation), and quantitative PCR to validate MeDIP array results by accurately measuring enrichment of methylated DNA fragments. |
|
| Deliverables |
| Can I get the images for my data? |
Yes, we can supply you with the raw data array images (.tif) upon request. Scaled log2-ratio data (.gff) files and peak (.gff) files are included in your deliverable data. |
| Can I get a graphical representation of all probes for a cataloged design that I am intersted in so I can see what regions of the genome have coverage? |
Yes, we can generate this information in GFF format for all of the catalog designs. You will need a copy of Roche NimbleGen's SignalMap software to view the GFF files. A free, 30-day demo version of SignalMap is available for download. |
| Does NimbleGen generate reports listing the most significant binding/modification events for my DNA methylation experiment? |
Yes, we are including two summary reports that map the peaks from your DNA methylation data relative to the transcription start site of a gene. For instance, if a peak is called within the promoter region of a gene, the report will give the approximate location of the peak as a negative position (upstream of the start site) or positive position (downstream). Also included are accession number of the gene, gene ID, chromosome position, among others. These reports will help narrow down the genomic regions you should be looking at when moving forward to validate your DNA methylation data (e.g. bisulfite sequencing or qPCR). |
|
| Benefits |
| Why should I use a genome-wide approach to study DNA methylation? |
NimbleGen's DNA methylation analysis arrays allow you to map DNA methylation across your genome of interest genome in an unbiased fashion, compare differential methylation patterns between cells, tissues, and tumors, and identify prognostic markers. |
| How sensitive are NimbleGen's DNA methylation arrays? |
Our arrays, when coupled with MeDIP, can readily detect as little as 2 methylated CpG dinucleotides per fragment. |
| What is the reproducibility of NimbleGen DNA methylation analysis arrays? |
While not fully validated by NimbleGen scientists, Weber et al. (Nat. Genet. 2007, 39(4):457-466) observed highly reproducible DNA methylation data between biological replicates (R = 0.92 for WI38 cells and R = 0.95 and 0.91 for sperm) when using MeDIP coupled with NimbleGen human promoter arrays. |
|
|
|
|
|
