Supervised mode


With supervised mode, difference across groups will be reported. We sugguest sorting or filtering DMRs by their length and absolute mean difference.

python /path_to_metilene3/metilene3.py [-i <string>] [-g <string>] [-o <string>] [optional options...]

Parameters:

(Go to DMR calling for more details!)

parameter unit default description
-i, --input string   the input methylation data, format
-g, --groupinfo string   the input group information table, format
-o, --output string   the output directory
-t, --threads integer 1 (optional) number of threads
-s, --seed integer 1 (optional) set seed for random generator
-O, --outputImputed bool False (optional) True or False, save the CpG methylation matrix with imputed values as file imputed.tsv
-p, --verbose bool False (optional) True or False, track the process
-M, --maxdist integer 300 (optional) maximum distance between two CpG, details
-m, --minCpGs integer 10 (optional) minimum CpGs, details
-d, --minMethDiff double 0.1 (optional) minimum mean methylation difference, details
-D, --minMethDiffHigh double 0.5 (optional) minimum mean methylation difference for GSEA, similar to -d, --minMethDiff but a higher value will be recommanded to reduce the number of false positive DMRs, details
-r, --minDMR integer 5 (optional) minimum CpGs with minimum mean methylation difference in a segment, details
-X, --minNonNA integer 1 (optional) minimum samples with non-NA values in each group
-v, --valley double 0.7 (optional) a cutoff for the difference between global and regional methylation differences, details
-k, --anova double 0.01 (optional) Kruskal-Wallis-Test p-value cutoff
-anno, --annotation string   (optional) hg19 or hg38, use ChIPseeker to annotate the DMRs
-refs, --refSeq string   (optional) reference genome, fasta file, for sequence annotation
-gsea, --genesets string   (optional) geneset gmt file for GSEA
-pdrl, --pandarallel bool True (optional) True or False, run Kruskal-Wallis-Test with pandarallel