PART 1: preparation before simulating

1. Input

   • hg19.fa.gz : The reference genome file
   • DMRs_unDMRs_signal.bed : The first three columns are location information and the fourth column represents the regions' classification(0: non-DMRs 2:hyper-DMRs 3:hypo-DMRs)
   • real_data_mean_chr21.bed: The mean CpG level of CpG sites from real data.

2. Output

   • test.fa: A folder that contains the fasta file for treatment group.
   • control.fa : A folder that contains the fasta file for control group.
   • test.bed: A folder that contains the fasta file for treatment group.
   • control.bed: A folder that contains the fasta file for control group.
   • total DMRs: The first three columns are the location information and the fourth column represents the regions' classification(0: non-DMRs 1:DMRs).

3. Script

   • preparation.sh

PART 2: The process of simulation

1. Input

   • control.fa : A folder that contains the fasta file for control group.
   • test.fa : A folder that contains the fasta file for treatment group.

2. Output

   • rrbssim: A folder that contains all simulated files.
   • fastqc_initial: A folder that contains the html files before trimming.
   • fq: A folder that contains the simulated fastq files.
   • fastqc_trim： A folder that contains the file after trimming.
   • bam: A folder that contains the bam files which were processed by bismark.
   • cov: A folder that contains the coverage files which were processed by bismark
   • bed: A folder that contains the bed files which generated from cov file.

3. Script

   • simulation.sh : We could get the 1 test simulation data and 1 control simulation data through running this script
     • The basic format of run this script is like: sh simulation.sh -o ${outdir} -t ${time}
   • run_simulation.sh : We could get n pairs of simulated data through running this script and this script is a supplementary for simulation.sh
     • The basic format of run this script is like: sh run_simulation.sh -o ${outdir} -n ${number}

PART 3: The process of calculating the overlap in simulated data

1. Input

   • *_DMR: The results of different tools about DMRs.
   • chr21.bed: The bed file of the start and end of chromosome21.
   • total DMRs: The first three columns are the location information and the fourth column represents the regions' classification(0: non-DMRs 1:DMRs).

2. Output

   • *_rate.bed: The overlap rate of diifferent tools.

3. Script

   • simulateddata_overlap.sh: The main script of calculating overlap rate in simulated data.

   • simulattedata_overlap.py: The sub script of calculating overlap rate in simulated data.

PART 4: The process of calculating the AUC

1. Input

   • *_DMR: The results of different tools about DMRs.
   • merge_total.bed: The bed file that contains the depth information of the intersection of CpG sites of all samples.
   • total DMRs: The first three columns are the location information and the fourth column represents the regions' classification(0: non-DMRs 1:DMRs).
   • chr21.bed: The bed file of the start and end of chromosome21.

2. Output

   • *_rate.bed: The overlap rate of diifferent tools.
   • AUC_results: The file that contains the value of AUC.

3. Script

   • calculate_AUC.sh: The main script of calculating AUC
   • simulationdata_overlap.py: The sub script that used to calculated the overlapped rate in simulated data.
   • calculate_AUC_all.R The sub script that used to calculate the AUC.PART 5: The process of calculating the AUC

PART 5: The process of calculating the kendall'W

1. Input

   • *_rate.bed: The overlap rate of diifferent tools.

2. Output

   • kendallW_results: The kendall‘s coefficient of concordance
   • MDS_results:The results of multidimensional scaling.

3. Script

   • kendall'W.R: The main script of calculating kendall'W.