DNA Methylation Data Analysis
4 likes3,678 views
This document provides an overview of DNA methylation analysis. It begins with background on DNA methylation functions and diseases. It then discusses methods for measuring DNA methylation status, including bisulfite sequencing. The document reviews steps for DNA methylation data analysis using tools like methylKit in R. It presents a case study example of analyzing DNA methylation data from human stem cells and fibroblasts. Alignment, quality control, differential methylation analysis and visualization are discussed.
![Mapping BS-Seq
FASTQ Files
(Example ONLY)
> export AdapterTrich=AATGATACGGCGACCACCGAGATCTACACTCTTTCCCTACACGACGCTCTTCCGATCT
> export AdapterArich=CAAGCAGAAGACGGCATACGAGCTCTTCCGATCT
> bsub -q 4G -o rmapbs.stdout -e rmapbs.stderr "
/work3/LSLNGSDNAMETH/methpipe-3.3.1/bin/rmapbs-pe
-c /work3/LSLNGSDNAMETH/methpipe-data/data/genome
-o /work3/USERNAME/Output/test.mr
-m 3 -L 400 -C $AdapterTrich:$AdapterArich
/work3/LSLNGSDNAMETH/methpipe-data/data/snippet_1.fq
/work3/LSLNGSDNAMETH/methpipe-data/data/snippet_2.fq"
37
> /work3/LSLNGSDNAMETH/methpipe-3.3.1/bin/rmapbs-pe
Usage: rmapbs-pe [OPTIONS] <fastq-reads-file>
Options:
-o, -output output file name
-c, -chrom chromosomes in FASTA file or dir
-T, -start index of first read to map
-N, -number number of reads to map
-s, -suffix suffix of chrom files (assumes dir provided)
-m, -mismatch maximum allowed mismatches
-M, -max-map maximum allowed mappings for a read
-C, -clip clip the specified adaptor
-L, -fraglen max fragment length
-suffix-len Suffix length of reads name
-v, -verbose print more run info
Help options:
-?, -help print this help message
-about print about message](https://image.slidesharecdn.com/20151118-academia-sinica-lsl-ngs-workshop-151116082244-lva1-app6892/85/DNA-Methylation-Data-Analysis-37-320.jpg)
![Differential methylated region (DMR)
> bsub -q 16G -o stdout -e stderr "
dmr /work3/LSLNGSDNAMETH/h1_imr90.chrX.methdiff
/work3/LSLNGSDNAMETH/h1.chrX.hmr /work3/LSLNGSDNAMETH/imr90.chrX.hmr
h1_lt_imr90.chrX.dmr h1_gt_imr90.chrX.dmr"
45
==> h1_lt_imr90.chrX.dmr <==
chrX 2727656 2728600 X:18 10 +
chrX 2731108 2731952 X:15 4 +
chrX 2732390 2733303 X:37 8 +
chrX 2740632 2740962 X:9 0 +
chrX 2758131 2758153 X:3 0 +
chrX 2817685 2817980 X:9 0 +
chrX 2855757 2855890 X:1 1 +
chrX 2890571 2890884 X:9 4 +
chrX 3004371 3004626 X:9 0 +
chrX 3238227 3238677 X:24 0 +
==> h1_gt_imr90.chrX.dmr <==
chrX 2825454 2826947 X:37 17 +
chrX 2857708 2857760 X:2 0 +
chrX 3272822 3273033 X:13 3 +
chrX 3275527 3275594 X:1 0 +
chrX 3287038 3289160 X:36 9 +
chrX 3643168 3643374 X:7 0 +
chrX 4016033 4022054 X:47 29 +
chrX 4028369 4042000 X:79 54 +
chrX 4051286 4059878 X:52 39 +
chrX 4079778 4087714 X:45 26 +
Number of significant differential methylated CpG
Meth. level lower in H1 than IMR90 Meth. level lower in IMR90 than H1
# of CpG
> awk -F "[:t]" ’$5 >= 10 && $6 >= 5 {print $0}’ h1_lt_imr90.chrX.dmr
> h1_lt_imr90.chrX.dmr.filtered](https://image.slidesharecdn.com/20151118-academia-sinica-lsl-ngs-workshop-151116082244-lva1-app6892/85/DNA-Methylation-Data-Analysis-45-320.jpg)
![Get descriptive stats on methylation
> png("test1.png",width=600,height=600)
> getMethylationStats(myobj[[1]],plot=T,both.strands=F)
> dev.off()
null device 1
> png("control1.png",width=600,height=600)
> getMethylationStats(myobj[[3]],plot=T,both.strands=F)
> dev.off()
null device 1
51](https://image.slidesharecdn.com/20151118-academia-sinica-lsl-ngs-workshop-151116082244-lva1-app6892/85/DNA-Methylation-Data-Analysis-51-320.jpg)
DNA Methylation Data Analysis
- 1. - Academia Sinica LSL NGS Workshop - DNA Methylation Data Analysis Yi-Feng Chang Ph.D. Molecular Medicine Research Center, Chang Gung University ianyfchang@mail.cgu.edu.tw 03-2118800 #3166 or #3528 2015/11/18 1
- 2. Outlines • DNA Methylation: Functions and Diseases • Methods of Measuring DNA Methylation Status • DNA Methylation Data Analysis • A Case Study of DNA Methylation Data Analysis • DNA Methylation Data Visualization 2
- 7. DNA Demethylation Pathway 7 Moore, L.D., Le, T. & Fan, G. DNA methylation and its basic function. Neuropsychopharmacology 38, 23-38 (2013). • 5mC: 5-Methylcytosine • 5hmC: 5-hydroxymethylcytosine • 5hmU: 5-hydroxymethyluracil • 5fC: 5-formylcytosine • 5caC: 5-carboxycytosine • Tet: Ten-eleven translocation enzymes • AID/ APOBEC: activation-induced cytidine deaminase/apolipo-protein B mRNA- editing enzyme complex • TDG: Thymine DNA glycosylase • SMUG1: Single-strand-selective monofunctional uracil-DNA glycosylase 1
- 9. Timeline of Technologies for Studying DNA Methylation 9 COBRA: Combined Bisulfite Restriction Analysis AP-PCR: Methylation-Sensitive Arbitrarily Primed PCR AIMS: DNA methylation by amplification of intermethylated sites RRBS: Reduced representation bisulfite sequencing MS-HRM: Methylation-sensitive high resolution melting MeDIP-Seq: Methylated DNA immunoprecipitation sequencing MethylC-Seq/BS-Seq: Bisulfite sequencing TAB-Seq: Tet-Assisted Bs-Seq MAB-Seq: M.SssI methylase-assisted BS-Seq MS-HRM MeDIP-Seq BS-Seq MethylC-Seq TAB-Seq MAB-Seq Harrison, A. & Parle-McDermott, A. DNA methylation: a timeline of methods and applications. Front Genet 2, 74 (2011). 2015
- 11. 11 Lister, R. & Ecker, J.R. Finding the fifth base: genome-wide sequencing of cytosine methylation. Genome Res 19, 959-66 (2009). Genomic DNA Deep Sequencing Techniques for Genome- Wide Sequencing of Cytosine Methylation Sites
- 12. 12 Genomic DNA Deep Sequencing Techniques for Enrichment of Methylated or Target Regions Prior to BS-Seq Lister, R. & Ecker, J.R. Finding the fifth base: genome-wide sequencing of cytosine methylation. Genome Res 19, 959-66 (2009).
- 13. Approaches for Detecting Active DNA Demethylation at Single Base Resolution 13 TAB-Seq: Tet-Assisted Bs-Seq Yu, M. et al. Tet-assisted bisulfite sequencing of 5- hydroxymethylcytosine. Nat Protoc 7, 2159-70 (2012). Yu, M. et al. Base-resolution analysis of 5- hydroxymethylcytosine in the mammalian genome. Cell 149, 1368-80 (2012). MAB-Seq: M.SssI methylase-assisted BS-Seq Wu, H., Wu, X., Shen, L. & Zhang, Y. Single-base resolution analysis of active DNA demethylation using methylase-assisted bisulfite sequencing. Nat Biotechnol 32, 1231-40 (2014).
- 14. Key Metrics of the Technology Comparison 14 Beck, S. Taking the measure of the methylome. Nat Biotechnol 28, 1026-8 (2010). Human Methylation 450K contains approximately 480k CpG sites, covering 99% RefSeq genes (hg19) and 96% CpG islands (CGIs).
- 15. Genomic Coverage of MeDIP-seq, MethylCap-seq, RRBS and Infinium 15 Bock, C. et al. Quantitative comparison of genome-wide DNA methylation mapping technologies. Nat Biotechnol 28, 1106-14 (2010). MeDIP-seq and MethylCap-seq provide broad coverage of the genome, whereas RRBS and Infinium are more restricted to CpG islands and promoter regions
- 17. WGBS Coverage Depth vs Replicates • Using several high-coverage reference data sets to experimentally determine minimal sequencing requirements 17 Ziller, M. J., Hansen, K. D., Meissner, A. & Aryee, M. J. Coverage recommendations for methylation analysis by whole-genome bisulfite sequencing. Nat Methods 12, 230-232, 231 p following 232, doi:10.1038/nmeth.3152 (2015).
- 18. WGBS Coverage Depth vs Replicates • For DMR identification • Per-sample coverage in the range of 5–15×, depending on the magnitude of methylation differences between the groups and whether a smoothing or single CpG-based DMR identification strategy is used • To identify long DMRs with large methylation differences, we find that reducing coverage down to 1× or 2× per sample is acceptable • Biological replicates should be analyzed separately to increase power, as opposed to being pooled together for analysis • Strongly argue for the use of at least two separate biological replicates for DMR analysis • Choosing an appropriate number of biological replicates is a complex issue influenced by the degree of within-group heterogeneity, the magnitude of between-group differences and the presence of confounding factors such as batch effects. 18Ziller, M. J., Hansen, K. D., Meissner, A. & Aryee, M. J. Coverage recommendations for methylation analysis by whole-genome bisulfite sequencing. Nat Methods 12, 230-232, 231 p following 232, doi:10.1038/nmeth.3152 (2015).
- 20. Effect and Problems of Bisulfite Treatment of DNA 20 Krueger, F., Kreck, B., Franke, A. & Andrews, S.R. DNA methylome analysis using short bisulfite sequencing data. Nat Methods 9, 145-51 (2012). Mapping bisulfite reads to 4 possible bisulfite strands (OT/CTOT/OB/CTOB) is equivalent to mapping the bisulfite read and its reverse complementary read to both Top/Bottom strands of the original reference sequence. OT, original top strand; CTOT, strand complementary to the original top strand; OB, original bottom strand; and CTOB, strand complementary to the original bottom strand.
- 21. How to Align BS Reads Against Reference Genome? 21 Bock, C. Analysing and interpreting DNA methylation data. Nat Rev Genet 13, 705-19 (2012) TCGA TCGT ACGT ATGA TTGT ATGTTCGA ATGA BS-Seq reads
- 22. Procedure to Perform Three-Letter Alignment 22 Krueger, F. & Andrews, S.R. Bismark: A flexible aligner and methylation caller for Bisulfite-Seq applications. Bioinformatics (2011).
- 23. Three-Letter Alignment 23 Multiple hits Bock, C. Analysing and interpreting DNA methylation data. Nat Rev Genet 13, 705-19 (2012)
- 24. Wild-Card Alignment 24 Convert C/T to Y Multiple hits Bock, C. Analysing and interpreting DNA methylation data. Nat Rev Genet 13, 705-19 (2012)
- 26. Workflow for Analyzing BS-Seq data 26Krueger, F., Kreck, B., Franke, A. & Andrews, S.R. DNA methylome analysis using short bisulfite sequencing data. Nat Methods 9, 145-51 (2012). http://omictools.com/bisulfite-seq/
- 29. Required Software in Your Laptop • Mac OS X Terminal • Application à Utilities à Terminal (終端機) • Linux console • Putty: http://the.earth.li/~sgtatham/putty/latest/x86/putty.exe • SCP/SFTP/FTP client • Winscp: http://winscp.net/download/winscp556.zip • PDF viewer • http://get.adobe.com/tw/reader/ • R • https://cran.r-project.org/ 29
- 30. Required R Packages • Bioconductor • http://www.bioconductor.org/install/#install- bioconductor-packages • methylKit: • https://github.com/al2na/methylKit 30 > R # dependencies > install.packages( c("data.table","devtools")) > source("http://bioconductor.org/biocLite.R") > biocLite(c("GenomicRanges","IRanges")) # install the development version from github > library(devtools) > install_github("al2na/methylKit",build_vignettes=FALSE)
- 31. Analysis Pipeline 31 Allele-specific Methylated Regions amrfinder allelicmeth Differential Methylation Region dmr Large Hypo/Hyper-Methylation Domains pmd Hypo/Hyper-Methylation Regions hmr hyperhmr pmr Methylation Calling methcounts Bisulfite Conversion Rate bsrate Remove Duplicate Reads duplicate-remover Mapping walt Quality Trimming fastq_masker Cross-species Comparison of Methylomes liftOver Calculating Methylation Ratio for Regions bigWigAverageOverBed roimethstat bwtools Generate Methylation BED file Bedtools bedGraphToBigWig fastx toolkit: http://hannonlab.cshl.edu/fastx_toolkit/ MethPipe: http://smithlabresearch.org/software/methpipe/ Bedtools: https://github.com/arq5x/bedtools2 Programs from UCSC Genome Browser: http://hgdownload.cse.ucsc.edu/admin/exe/linux.x86_64 bwtool: https://github.com/CRG-Barcelona/bwtool/wiki Sorting mr files Sorting mr files http://smithlabresearch.org/downloads/methpipe-manual.pdf
- 32. Public BS-Seq Datasets 32 http://smithlabresearch.org/software/methbase/ Other species in NCBI GEO Database • Glycine max (Soy beans) • Schistocerca gregaria (Locust) • Rattus norvegicus (Rat) • Danio rerio (Zebra fish) • Drosophila melanogaster (Fruit fly) • Oryza sativa (Rice) • Macaca mulatta (Rhesus monkey) • Mus musculus domesticus (Western Europen house mouse) • Xenopus (Silurana) tropicalis (Frog) • Cynoglossus semilaevis (Tongue sole, bony fish) • Bombyx mori (Silkworm) • Harpegnathos saltator (Jerdon's jumping ant) • Camponotus floridanus (Florida carpenter ant)
- 34. Convert SRA to FASTQ (Example ONLY) # sra-toolkit can be download from https://github.com/ncbi/sratoolkit > fastq-dump --split-3 SRR018975.sra > ls SRR018975.fastq 34
- 35. DEMO Files > cd /work3/LSLNGSDNAMETH > ls -alh total 12G drwxr-xr-x 4 u00gel00 u00ycm02 4.0K Nov 16 00:29 . drwxrwxrwt 109 root root 4.0K Nov 15 14:10 .. -rwxr-xr-x 1 u00gel00 u00ycm02 65K Nov 15 17:22 h1.chrX.hmr -rwxr-xr-x 1 u00gel00 u00ycm02 4.6G Nov 15 14:51 h1.chrX.mr.dremove -rwxr-xr-x 1 u00gel00 u00ycm02 9.8K Nov 15 17:22 h1.chrX.pmd -rwxr-xr-x 1 u00gel00 u00ycm02 34M Nov 15 17:39 h1.chrX_CpG.meth -rwxr-xr-x 1 u00gel00 u00ycm02 39M Nov 15 23:52 h1.chrX_CpG.meth.for.methylKit -rwxr-xr-x 1 u00gel00 u00ycm02 161K Nov 15 17:22 h1_gt_imr90.chrX.dmr -rwxr-xr-x 1 u00gel00 u00ycm02 45M Nov 15 17:22 h1_imr90.chrX.methdiff -rwxr-xr-x 1 u00gel00 u00ycm02 55K Nov 15 17:22 h1_lt_imr90.chrX.dmr -rwxr-xr-x 1 u00gel00 u00ycm02 194K Nov 15 17:22 imr90.chrX.hmr -rwxr-xr-x 1 u00gel00 u00ycm02 7.3G Nov 15 14:52 imr90.chrX.mr.dremove -rwxr-xr-x 1 u00gel00 u00ycm02 5.6K Nov 15 17:22 imr90.chrX.pmd -rwxr-xr-x 1 u00gel00 u00ycm02 35M Nov 15 17:39 imr90.chrX_CpG.meth -rwxr-xr-x 1 u00gel00 u00ycm02 40M Nov 15 23:52 imr90.chrX_CpG.meth.for.methylKit drwxr-xr-x 6 u00gel00 u00ycm02 4.0K Nov 15 14:28 methpipe-3.3.1 drwxr-xr-x 4 u00gel00 u00ycm02 4.0K Nov 15 14:46 methpipe-data 35
- 36. Quality Trimming and Split FASTQ Files into Smaller Files (Example ONLY) #e.g. SRR018975.fastq.gz > for f in *.gz; do b=`basename $f .gz`; echo $f bsub -q 4G -o $f.stdout -e $f.stderr " gzip -dc $f| fastq_masker -q 30 -Q33| split -dl 6000000 - $b- "; done > ls SRR018975.fastq-00 SRR018975.fastq-01 SRR018975.fastq-02 … 36 #e.g. SRR018975.fastq.gz # listing all gzip files one by one # SRR018975.fastq #uncompressing gzip file and out to stdout #masking low quality reads as Ns #spliting fastq file into smaller ones
- 37. Mapping BS-Seq FASTQ Files (Example ONLY) > export AdapterTrich=AATGATACGGCGACCACCGAGATCTACACTCTTTCCCTACACGACGCTCTTCCGATCT > export AdapterArich=CAAGCAGAAGACGGCATACGAGCTCTTCCGATCT > bsub -q 4G -o rmapbs.stdout -e rmapbs.stderr " /work3/LSLNGSDNAMETH/methpipe-3.3.1/bin/rmapbs-pe -c /work3/LSLNGSDNAMETH/methpipe-data/data/genome -o /work3/USERNAME/Output/test.mr -m 3 -L 400 -C $AdapterTrich:$AdapterArich /work3/LSLNGSDNAMETH/methpipe-data/data/snippet_1.fq /work3/LSLNGSDNAMETH/methpipe-data/data/snippet_2.fq" 37 > /work3/LSLNGSDNAMETH/methpipe-3.3.1/bin/rmapbs-pe Usage: rmapbs-pe [OPTIONS] <fastq-reads-file> Options: -o, -output output file name -c, -chrom chromosomes in FASTA file or dir -T, -start index of first read to map -N, -number number of reads to map -s, -suffix suffix of chrom files (assumes dir provided) -m, -mismatch maximum allowed mismatches -M, -max-map maximum allowed mappings for a read -C, -clip clip the specified adaptor -L, -fraglen max fragment length -suffix-len Suffix length of reads name -v, -verbose print more run info Help options: -?, -help print this help message -about print about message
- 38. Example Output of imr90 chrX 38 > head -n 30 /work3/LSLNGSDNAMETH/imr90.chrX.mr.dremove |column MR Format •RNAME (chromosome name) •SPOS (start position, 0-based) •EPOS (end position, 0-based) •QNAME (read name) •MISMATCH (number of mismatches) •STRAND (forward or reverse strand) •SEQ •QUAL
- 39. Remove Duplicates (Example ONLY) > export PATH=$PATH:/pkg/biology/methpipe/methpipe-3.3.1/bin/ > bsub -q 16G -o stdout -e stderr " LC_ALL=C sort -S 14G -k 1,1 -k 2,2n -k 3,3n -k 6,6 -o /work3/USERNAME/h1.chrX.mr.sorted_start /work3/LSLNGSDNAMETH/h1.chrX.mr; duplicate-remover -S /work3/USERNAME/h1.chrX_dremove_stat.txt -o /work3/USERNAME/h1.chrX.mr.dremove /work3/USERNAME/h1.chrX.mr.sorted_start " > cat stdout Successfully completed. Resource usage summary: CPU time : 343.80 sec. Max Processes : 3 Max Threads : 4 39 > cat/work3/USERNAME/h1.chrX_dremove_stat.txt TOTAL READS IN: 24350707 GOOD BASES IN: 1987943796 TOTAL READS OUT: 22884736 GOOD BASES OUT: 1867152730 DUPLICATES REMOVED: 1465971 READS WITH DUPLICATES: 1219174
- 40. Computing single-site methylation levels (Example Only) # sorting again for methylated CpG analysis bsub -q 16G -o stdout -e stderr " LC_ALL=C sort -S 14G -k 1,1 -k 3,3n -k 2,2n -k 6,6 -o /work3/USERNAME/h1.chrX.mr.sorted_end_first /work3/LSLNGSDNAMETH/h1.chrX.mr.dremove" # methylation calling bsub -q 16G -o stdout -e stderr " methcounts -c /work3/LSLNGSDNAMETH/hg18 -o /work3/USERNAME/h1.chrX.meth /work3/USERNAME/h1.chrX.mr.sorted_end_first" #extract CpG sites bsub -q 16G -o stdout -e stderr " symmetric-cpgs -o /work3/USERNAME/h1.chrX_CpG.meth h1.chrX.meth" 40 chrX 152 + CpG 0 0 chrX 232 + CpG 0 0 chrX 330 + CpG 0 0 chrX 334 + CpG 0 0 chrX 336 + CpG 0 0 chrX 364 + CpG 0 0 chrX 366 + CpG 0 0 chrX 374 + CpG 0 0 chrX 376 + CpG 0 0 meth ratio read count
- 41. Computation of methylation level statistics (Example ONLY) 41 bsub -q 16G -o stdout -e stderr " levels -o /work3/USERNAME/Output/h1.chrX.levels /work3/USERNAME/h1.chrX.meth"
- 42. Estimating bisulfite conversion rate > bsub -q 16G -o stdout -e stderr " bsrate -c /work3/LSLNGSDNAMETH/hg18 -o /work3/USERNAME/Output/h1.chrX.bsrate /work3/LSLNGSDNAMETH/h1.chrX.mr.dremove" 42 # head –n 16 /work3/USERNAME/Output/h1.chrX.bsrate OVERALL CONVERSION RATE = 0.980192 POS CONVERSION RATE = 0.980204 96942555 NEG CONVERSION RATE = 0.980179 96821402 BASE PTOT PCONV PRATE NTOT NCONV NRATE BTHTOT BTHCONV BTHRATE ERR ALL ERRRATE 1 1798190 1762518 0.98016 1796291 1760655 0.98016 3594481 3523173 0.98016 36327 3630808 0.01001 2 1654252 1617801 0.97797 1649805 1613025 0.97771 3304057 3230826 0.97784 41299 3345356 0.01235 3 1646403 1615036 0.98095 1644710 1613525 0.98104 3291113 3228561 0.98099 48231 3339344 0.01444 4 1699787 1666286 0.98029 1695105 1662078 0.98052 3394892 3328364 0.98040 50697 3445589 0.01471 5 1663363 1631006 0.98055 1658397 1626045 0.98049 3321760 3257051 0.98052 52464 3374224 0.01555 6 1720978 1687130 0.98033 1716036 1682351 0.98037 3437014 3369481 0.98035 45366 3482380 0.01303 7 1677561 1644979 0.98058 1677119 1644343 0.98046 3354680 3289322 0.98052 53873 3408553 0.01581 8 1714426 1681206 0.98062 1714378 1681339 0.98073 3428804 3362545 0.98068 34491 3463295 0.00996 9 1702891 1668424 0.97976 1700092 1665742 0.97980 3402983 3334166 0.97978 34861 3437844 0.01014 10 1681522 1648092 0.98012 1680471 1647068 0.98012 3361993 3295160 0.98012 45776 3407769 0.01343 11 1664207 1631036 0.98007 1664386 1631083 0.97999 3328593 3262119 0.98003 46055 3374648 0.01365 12 1651326 1618334 0.98002 1649370 1616514 0.98008 3300696 3234848 0.98005 44139 3344835 0.01320
- 43. Hypomethylated (hmr) and hypermethylated (hypermr) > bsub -q 16G -o stdout -e stderr " hmr -o /work3/USERNAME/h1.chrX.hmr /work3/USERNAME/h1.chrX_CpG.meth" > bsub -q 16G -o stdout -e stderr " pmd -o /work3/USERNAME/h1.chrX.pmd /work3/USERNAME/h1.chrX_CpG.meth" 43 chrX 2727656 2728600 HYPO0 18 + chrX 2731108 2731952 HYPO1 14 + chrX 2732390 2733303 HYPO2 23 + chrX 2740632 2740962 HYPO3 9 + chrX 2756524 2758153 HYPO4 139 + chrX 2817685 2817980 HYPO5 8 + chrX 2855757 2857708 HYPO6 127 + chrX 2890571 2890884 HYPO7 9 + chrX 3004371 3004626 HYPO8 9 + chrX 3238227 3238677 HYPO9 9 + # of CpG
- 44. Differential Methylation Analysis > bsub -q 16G -o stdout -e stderr " methdiff -o /work3/USERNAME/h1_imr90.chrX.methdiff /work3/LSLNGSDNAMETH/h1.chrX_CpG.meth /work3/LSLNGSDNAMETH/imr90.chrX_CpG.meth" 44 chrX 2709681 + CpG 0.749276 7 2 12 7 chrX 2709727 + CpG 0.917633 4 1 9 12 chrX 2709774 + CpG 0.894737 3 1 6 10 chrX 2709871 + CpG 0.742424 0 16 0 48 chrX 2709890 + CpG 0.857575 3 20 3 47 chrX 2709982 + CpG 0.999354 10 2 7 19 chrX 2710014 + CpG 0.704043 3 6 3 10 chrX 2710023 + CpG 0.600782 4 3 4 4 chrX 2710146 + CpG 0.523077 1 2 8 14 chrX 2710155 + CpG 0.234026 3 3 17 9 Probability Sample A Un-meth Sample A Meth Sample B Un-meth Sample B Meth
- 45. Differential methylated region (DMR) > bsub -q 16G -o stdout -e stderr " dmr /work3/LSLNGSDNAMETH/h1_imr90.chrX.methdiff /work3/LSLNGSDNAMETH/h1.chrX.hmr /work3/LSLNGSDNAMETH/imr90.chrX.hmr h1_lt_imr90.chrX.dmr h1_gt_imr90.chrX.dmr" 45 ==> h1_lt_imr90.chrX.dmr <== chrX 2727656 2728600 X:18 10 + chrX 2731108 2731952 X:15 4 + chrX 2732390 2733303 X:37 8 + chrX 2740632 2740962 X:9 0 + chrX 2758131 2758153 X:3 0 + chrX 2817685 2817980 X:9 0 + chrX 2855757 2855890 X:1 1 + chrX 2890571 2890884 X:9 4 + chrX 3004371 3004626 X:9 0 + chrX 3238227 3238677 X:24 0 + ==> h1_gt_imr90.chrX.dmr <== chrX 2825454 2826947 X:37 17 + chrX 2857708 2857760 X:2 0 + chrX 3272822 3273033 X:13 3 + chrX 3275527 3275594 X:1 0 + chrX 3287038 3289160 X:36 9 + chrX 3643168 3643374 X:7 0 + chrX 4016033 4022054 X:47 29 + chrX 4028369 4042000 X:79 54 + chrX 4051286 4059878 X:52 39 + chrX 4079778 4087714 X:45 26 + Number of significant differential methylated CpG Meth. level lower in H1 than IMR90 Meth. level lower in IMR90 than H1 # of CpG > awk -F "[:t]" ’$5 >= 10 && $6 >= 5 {print $0}’ h1_lt_imr90.chrX.dmr > h1_lt_imr90.chrX.dmr.filtered
- 46. Other Utilities • DM analysis of two groups of DNA methylomes • Robinson, M. D. et al. Statistical methods for detecting differentially methylated loci and regions. Frontiers in genetics 5, 324, doi:10.3389/fgene.2014.00324 (2014). • Allele-specific methylation • allelicmeth • amrfinder: http://smithlabresearch.org/software/amrfinder/ • Estimate hydroxymethylation(5hmC) and methylation (5mC) levels from BS-seq, oxBS-seq and TAB-seq • mlml: http://smithlabresearch.org/software/mlml/ 46
- 48. R Packages: methylKit The following examples were adopt from the tutorials of methylKit • Akalin, A. et al. methylKit: a comprehensive R package for the analysis of genome-wide DNA methylation profiles. Genome Biol 13, R87, doi:10.1186/gb-2012-13-10-r87 (2012). • Tutorial: http://methylkit.googlecode.com/files/methylKitTutorial_feb2012. pdf • Tutorial Slide: http://methylkit.googlecode.com/files/ methylKitTutorialSlides_2013.pdf 48
- 49. Convert MethPipe mr Format to methylKit Format Id chr base strand coverage freqC freqT Chr21.9764539 chr21 9764539 R 12 25.00 75.00 Chr21.9764513 chr21 9764513 R 12 0.00 100.00 Chr21.9820622 chr21 9820622 F 13 0.00 100.00 Chr21.9837545 chr21 9837545 F 11 0.00 100.00 Chr21.9849022 chr21 9849022 F 124 72.58 27.42 Chr21.9853326 chr21 9853326 F 17 70.59 29.41 49 > awk -F $'t' -v OFS=$'t’ '$6>0{$5=int($5*100); print $1"."$2, $1, $2, "F", $6, $5, (100-$5)}' /work3/LSLNGSDNAMETH/h1.chrX_CpG.meth > /work3/USERNAME/Output/h1.chrX_CpG.meth.for.methylKit > awk -F $'t' -v OFS=$'t' '$6>0{$5=int($5*100); print $1"."$2, $1, $2, "F", $6, $5, (100-$5)}' /work3/LSLNGSDNAMETH/imr90.chrX_CpG.meth > /work3/USERNAME/Output/imr90.chrX_CpG.meth.for.methylKit
- 50. Read Methylation Files into methylKit Objects > library(methylKit) # load methylation files (change to your datasets) > file.list=list( system.file("extdata", "test1.myCpG.txt", package = "methylKit"), system.file("extdata", "test2.myCpG.txt", package = "methylKit"), system.file("extdata", "control1.myCpG.txt", package = "methylKit"), system.file("extdata", "control2.myCpG.txt", package = "methylKit") ) # read the files to a methylRawList object: myobj > myobj=read( file.list, sample.id=list("test1", "test2","ctrl1","ctrl2"), assembly="hg18",treatment=c(1,1,0,0)) > head(myobj) 50
- 51. Get descriptive stats on methylation > png("test1.png",width=600,height=600) > getMethylationStats(myobj[[1]],plot=T,both.strands=F) > dev.off() null device 1 > png("control1.png",width=600,height=600) > getMethylationStats(myobj[[3]],plot=T,both.strands=F) > dev.off() null device 1 51
- 52. Sample Correlation > png("correlation.png",width=1000,height=1000) > getCorrelation(meth, plot = T) test1 test2 ctrl1 ctrl2 test1 1.0000000 0.9252530 0.8767865 0.8737509 test2 0.9252530 1.0000000 0.8791864 0.8801669 ctrl1 0.8767865 0.8791864 1.0000000 0.9465369 ctrl2 0.8737509 0.8801669 0.9465369 1.0000000 > dev.off() 52
- 53. Get bases covered by all samples and cluster samples # merge all samples to one table by using base-pair locations that are covered in all samples > meth=unite(myobj) # cluster all samples using correlation distance and plot hierarchical clustering > png("cluster.png", width=600, height=600) > hc = clusterSamples(meth, dist="correlation", method="ward", plot=T) > dev.off() > png("pca.png", width=600,height=600) > PCASamples(meth) > dev.off() 53
- 54. Calculate differential methylation # calculate differential methylation p-values and q-values > myDiff=calculateDiffMeth(meth) # get differentially methylated regions with 25% difference and qvalue < 0.01 > myDiff25p=get.methylDiff(myDiff,difference=25,qvalue=0.01) # get differentially hypo methylated regions with 25% difference and qvalue<0.01 > myDiff25pHypo =get.methylDiff(myDiff,difference=25,qvalue=0.01,type="hypo") # get differentially hyper methylated regions with 25% difference and qvalue<0.01 > myDiff25pHyper=get.methylDiff(myDiff,difference=25,qvalue=0.01,type="hyper") 54
- 55. Differential methylation events per chromosome > png("meth_event.png",width=600,height=600) > diffMethPerChr(myDiff, plot = T, qvalue.cutoff = 0.01,meth.cutoff = 25) > dev.off() 55
- 57. Annotating Differential Methylation Events around CpG Islands > cpg.obj = read.feature.flank(system.file("extdata", "cpgi.hg18.bed.txt", package = "methylKit"),feature.flank.name = c("CpGi", "shores")) > diffCpGann = annotate.WithFeature.Flank(myDiff25p,cpg.obj$CpGi, cpg.obj$shores, feature.name = "CpGi",flank.name = "shores") 57
- 59. Questions? 59