📣 DNA methylation heterogeneity measures the epigenetic diversity within a cell population where the behaviour of individual cells can vary as differential responses to the environmental stimuli or dynamic progression of cellular development.
MeH users guide is available as a PDF file, containing the detail of each step. For questions please open an issue on GitHub or contact me.
Please follow the tutorial of example use case.
- MeH 🐑
- python 2.7 +
- pandas package 0.24 +
- pysam package 0.16.0.1 +
- joblib package
- R 3.4 +
MeH can be installed for Linux, macOS, or Windows by either compiling from source which has the advantage that it will be optimized to the specific system:
git clone https://github.com/PaoyangLab/MeH.git
cd MeHUse the scrpit MeHscr.py to calculated the methylation heterogeneity.
❕used as command-line in your terminal.
- Run all the files under folder "MeHdata", including:
- .bam and .bam.bai files
- .fa and .fa.fai of the reference genome
it is recommend removing potential PCR duplicates for the .bam file by Samtools before input to MeH.
samtools rmdup <input.bam> <output.bam>
$ python MeHscr.py -h
usage: MeHscr.py [-h] [-w WINDOWSIZE] [-c CORES] [-m MEH] [-d DIST] [--CG]
[--CHG] [--CHH] [--opt] [--mlv] [--imp]
optional arguments:
-h, --help show this help message and exit
-w WINDOWSIZE, --windowsize WINDOWSIZE
number of CGs
-c CORES, --cores CORES
number of cores
-m MEH, --MeH MEH Methylation heterogeneity score 1:Abundance 2:PW
3:Phylogeny [Default: 2]
-d DIST, --dist DIST Distance between methylation patterns 1:Hamming 2:WDK [Default: 1]
--CG Include genomic context CG
--CHG Include genomic context CHG
--CHH Include genomic context CHH
--opt Outputs compositions of methylation patterns
--mlv Outputs methylation levels
--imp Implement BSImp (impute if valid)
# 'CG' only with window size of 4 cytosines and 4 cores parallel processing (default score is pairwise-similarity-based method, default distance between methylation patterns is Hamming distance)
python MeHscr.py -w 4 -c 4 --CG
# 'CG', 'CHG' and 'CHH' with window size of 4 cytosines, weighted degree kernel for pairwise distances between methylation patterns and 8 cores parallel processing
python MeHscr.py -w 4 -c 8 --CG --CHG --CHH -d 2
# 'CG', 'CHG' and 'CHH' with window size of 4 cytosines, output methylation pattern compositions within each window, all using 8 cores parallel processing and no imputation
python MeHscr.py -w 4 -c 8 --CG --CHG --CHH --opt --impThe programme is running at folder "/MeHdata"
- MeHscreening.log
Sample AT31test has coverage 5240 for context CG out of data coverage 192834
Sample AT33test has coverage 5236 for context CG out of data coverage 193431
Sample AT35test has coverage 5203 for context CG out of data coverage 192548
Sample AT37test has coverage 5233 for context CG out of data coverage 192694
- /MeHdata/sample.0.csv files for each sample
## CG_AT31test_0.csv in the example
chrom,pos,MeH,dis,strand
1,511,1.41421,139,f
1,791,2.7161,114,r
1,810,3.69631,102,r
1,840,4.11599,109,rFormat desctiptions:
(1) chromsome (2) position (3) Methlyation heterogeneity (4) distance between methylation patterns (5) strand as 'f' for forward or 'r' for reverse
- /MeHdata/Results.csv files for summary results
## CG_Results.csv in the example
chrom,bin,strand,AT31test,AT33test,AT37test,AT35test
1,600,f,1.41421,4.42434,1.97092,2.219035
1,600,r,2.7161,2.59751,3.62414,2.79942
1,1000,r,3.90615,4.90306,6.5213,4.0907849999999994
1,2600,r,0.0,0.707105,0.0,0.0Format desctiptions:
(1) chrom: chromsome (2) bin: position (of bin), specified to be at the centre of the bin; i.e. 600 means (400,800] (3) strand: f(orward)/r(everse) (4)-(6) Methlyation heterogeneity for each sample specified by the column header
Use the scrpit tobed.R to view results on IGV, use the code below to generate .bedGraph for all libraries.
❕used as command-line in your terminal.
- Results.csv files after calculating the methylation heterogeneity.
$ Rscript tobed.R -h
usage: tobed.R [--] [--help] [--opts OPTS] [-m -M [-r -R
MeH result file to .bedGraph
flags:
-h, --help show this help message and exit
optional arguments:
-x, --opts RDS file containing argument values
-m, input Meh resultes csv file
-r, reverse strand as negative MeH [default: all]# reverse strand as negative MeH
Rscript tobed.R -m ./MeHdata/CG_Results.csv
# reverse strand as MeH but starting at position = bin+1
Rscript tobed.R -m ./MeHdata/CG_Results.csv -r n- the .bedGraph for each sample and this will give you n filed with extension .bedGraph which can be opened using IGV.
Use the scrpit finddhr.R to find the differential hetergeneous regions (DHR).
❕used as command-line in your terminal.
- Results.csv files after calculating the methylation heterogeneity.
- genelist.txt with each row representing a gene and consists of gene name, chromosome ,strand, TSS,and TES.
$ Rscript finddhr.R -h
usage: finddhr.R [--] [--help] [--opts OPTS] [-m -M [-s -S [-g -G [-o
-O [-c -C [-p -P [-adjp -ADJP [-pvalue -PVALUE [-delta -DELTA
Find DHR
flags:
-h, --help show this help message and exit
optional arguments:
-x, --opts RDS file containing argument values
-m, input Meh resultes csv file
-s, Define conditions of all samples
-g, The gene list
-o, output gene list result csv file
-c, the number of core for analysis [default: 4]
-p, the region of promoter [default: 1000]
-adjp, Select differential heterogeneous regions based on adjust pvalue [default: 0.05]
-pvalue, Select differential heterogeneous regions based on pvalue [default: 0.05]
-delta, Select differential heterogeneous regions based on delta [default: 1.4]# An example is for W vs D with two replicates under specified conditions after t-statistics and propoter regions are 1000 bp.
Rscript finddhr.R -m ./MeHdata/CG_Results_test.csv -g ./MeHdata/genelist.txt -o ./MeHdata/CG -s W,W,D,D -p 1000- CG_DHR_Result.csv shows the list of DHR in down/up regulated gene/promoter
DHG Genebodys up: CHI3L1, DDX11L1, WASH7P, FAM138F, ATP1A1, GBP4
DHG Genebodys down:
DHG Promoter up: CHI3L1, DDX11L1, WASH7P, MIR1302-10, FAM138F, ATP1A1
DHG Promoter down:- CG_MeH_Result.csv shows the table with the differnece of MeH regions
chrom bin strand delta pvalue mean2 mean1 padj Gene Promoter
1 13000 f 10.000000 NaN 10.00000 0.00000000 NaN DDX11L1 NA
1 13800 f 10.000000 NaN 10.00000 0.00000000 NaN DDX11L1 NA
1 20200 f 9.528597 0.0314695032 10.00000 0.47140333 1.00000000 NA NA
1 21000 f 9.686856 0.0205726428 10.00000 0.31314395 1.00000000 NA NA
1 21000 r 9.175044 0.0081767503 10.00000 0.82495583 0.98938678 WASH7P NA
1 21400 f 10.000000 0.0011092571 10.23570 0.23570167 0.18524594 NA NA
1 21400 r 9.469671 0.0356153305 10.00000 0.53032875 1.00000000 WASH7P NAFormat desctiptions:
(1) chrom: chromsome (2) bin: position (of bin), specified to be at the centre of the bin; i.e. 600 means (400,800] (3) strand: f(orward)/r(everse) (4) dalta: mean of condition A - mean of condition B (5) pvalue: p-value after t-test (6) mean2: mean of condition B (7) mean1: mean of condition A (8) padj: adjust p-value (9) Gene: gene in this region (10) Promoter: promoter in this region
Use the scrpit IGV to find the differential hetergeneous regions (DHR).
- .bedGraph files after tobed.R
Pei-Yu Lin- GitHub