Scribd Logo
Upload

Welcome to Scribd!

  • 60 views

    Raharris - Illumina Infinium 450K Array

    Uploaded by

    keyeoh
    The document summarizes the Illumina Infinium 450K DNA methylation array. It describes the array design including over 487,000 probes assaying CpG sites, CpH sites, and SNPs. It discusses calculating beta values as a measure of methylation level and preprocessing steps like converting beta to M-values, normalizing between probe designs, and removing SNP-containing probes. Statistical methods for identifying differentially methylated regions (DMRs) are also summarized, including software like Illumina Methylation Analyzer and Limma.

    Copyright:

    Attribution Non-Commercial (BY-NC)
    Download as PDF, TXT or read online from Scribd

    Raharris - Illumina Infinium 450K Array

    Uploaded by

    keyeoh
    60 views10 pages
    The document summarizes the Illumina Infinium 450K DNA methylation array. It describes the array design including over 487,000 probes assaying CpG sites, CpH sites, and SNPs. It discusses calculating beta values as a measure of methylation level and preprocessing steps like converting beta to M-values, normalizing between probe designs, and removing SNP-containing probes. Statistical methods for identifying differentially methylated regions (DMRs) are also summarized, including software like Illumina Methylation Analyzer and Limma.

    Original Title

    Raharris.illumina Infinium 450K Array

    Copyright

    © Attribution Non-Commercial (BY-NC)

    Available Formats

    PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    The document summarizes the Illumina Infinium 450K DNA methylation array. It describes the array design including over 487,000 probes assaying CpG sites, CpH sites, and SNPs. It discusses calculating beta values as a measure of methylation level and preprocessing steps like converting beta to M-values, normalizing between probe designs, and removing SNP-containing probes. Statistical methods for identifying differentially methylated regions (DMRs) are also summarized, including software like Illumina Methylation Analyzer and Limma.

    Copyright:

    Attribution Non-Commercial (BY-NC)
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    60 views10 pages

    Raharris - Illumina Infinium 450K Array

    Uploaded by

    keyeoh
    The document summarizes the Illumina Infinium 450K DNA methylation array. It describes the array design including over 487,000 probes assaying CpG sites, CpH sites, and SNPs. It discusses calculating beta values as a measure of methylation level and preprocessing steps like converting beta to M-values, normalizing between probe designs, and removing SNP-containing probes. Statistical methods for identifying differentially methylated regions (DMRs) are also summarized, including software like Illumina Methylation Analyzer and Limma.

    Copyright:

    Attribution Non-Commercial (BY-NC)
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    You are on page 1of 10

    Illumina Infinium 450K Array

    Alan Harris
    Baylor College of Medicine Epigenomics Workshop

    Array Design
    487,557 probes assaying 12 samples
    CpG 482,421 CpH 3,091 methylated in embryonic stem cells rs SNPs 65

    Color of bead types: 350,076 (70%) Both (M,U) Design II 46,298 (10%) Green (M,U) Design I 89,203 (20%) Red (M,U) Design I

    Beta Values
    Beta value () - estimate of methylation level using ratio of intensities between methylated and unmethylated alleles
    = Methylated allele intensity (M) / (Unmethylated allele intensity (U) + Methylated allele intensity (M) + 100)

    Genome Studio Methylation Module Normalization


    Normalization to internal controls targeting same region in housekeeping genes with no CpG sites. Intensity multiplied by a constant normalization factor (for all samples) and divided by the average of normalization controls in the probes channel in the given sample Background subtraction derived by averaging the signals of built-in negative control probes

    High correlation with other bisulfite-based data1


    technical replicates - R2 > 0.992 27K BeadChip data - R2 > 0.95 (94% of 27K probes in 450K ) whole-genome bisulfite sequencing data - R2 > 0.950.96
    1Bibikova

    et al. (2011) Genomics 98:288

    Probe Annotations
    In Illumina Manifest Genomic Coordinates UCSC RefGene Name UCSC RefGene Accession UCSC RefGene Group UCSC CpG Islands Name Relation to UCSC CpG Island (Island, Shore, Shelf) Phantom DMR Enhancer HMM Island Regulatory Feature Name Regulatory Feature Group

    Preprocessing Convert Beta to M values?


    Comparison of Beta and M values1
    Relationship between Beta-value and M-value is a logit transformation Beta-value method has severe heteroscedasticity for highly methylated or unmethylated CpG sites M-value method provides much better performance in terms of detection rate and true positive rate for both highly methylated and unmethylated CpG sites Beta-value has a more intuitive biological interpretation, but the M-value is more statistically valid

    Software for Beta to M value conversion


    Lumi2 (R) Methylumi3 (R)

    1Du

    et al. (2010) BMC Bioinformatics. 11:587 2Du et al. (2008) Bioinformatics. 24:1547 3Davis et al. Bioconductor R package

    Preprocessing - Normalization
    Illumina claims probe design differences do not significantly affect differential methylation detection; can detect delta beta of 0.2 with 99% confidence1 Design I signals more stable and have an extended dynamic range of methylation values compared with design II signals2 Software to normalize between probe designs
    Illumina Methylation Analyzer (IMA)3 (R) peak correction Complete Pipeline4 (R) subset quantile normalization BMIQ5 (R) - beta-mixture quantile normalization

    1Bibikova

    et al. (2011) Genomics. 98:288 2Dedeurwaerder et al. (2011) Epigenomics. 3:771 3Wang et al. (2012) Bioinformatics. 28:729 4Touleimat et al. (2012) Epigenomics. 4:325 5Teschendorff et al. (2013) Bioinformatics. 29:189

    Preprocessing Remove SNPs


    SNPs in probes can lead to incorrect methylation measurements File of SNP containing probes can be downloaded from here: https://www.rforge.net/IMA/snpsites.txt 91988 cg probes contain SNPs

    Software to remove probes containing SNPs


    Illumina Methylation Analyzer (IMA)1 (R) Genboree Workbench Array Data Importer has option to exclude SNP containing probes

    1Wang

    et al. (2012) Bioinformatics. 28:729

    Differentially Methylated Regions


    Detection of statistically significant differentially methylated regions (DMRs) is primary analysis Multiple testing correction should be applied to statistical results A number of software packages have been developed to identify DMRs

    Illumina Methylation Analyzer (IMA)


    Calculates methylation indices for 5 UTR, first exon, gene body, 3 UTR, CpG island, CpG shore, CpG shelf Mean Median Tukeys Biweight robust average Identifies DMRs in regions Wilcoxon rank-sum test Students t-test Empirical Bayes Generalized linear models Multiple Testing Correction Bonferroni False Discovery Rate

    Limma
    Originally designed for detecting differential expression from arrays1

    Also widely used for Infinium methylation arrays


    Fits a linear model to the data for each gene Empirical Bayes method to moderate standard deviations between genes constraining the within-block correlations to be equal between genes Accessible through the Genboree Workbench

    1Smyth.

    (2004) Statistical Applications in Genetics and Molecular Biology. 3, No. 1,

    Article 3

    You might also like