CID - Chromatin Interaction Discovery for ChIA-PET and HiChIP data

Installation

CID:

• Java 1.7 or higher is required (available for Windows, Mac OS X, Linux, and Solaris): https://java.com/en/download/
• CID is part of the GEM software: download the GEM (version 3.4) JAR file at the GEM download page

Test installation by running: java -Xmx32G -jar gem.jar CID -h

MICC:

• R 3.3.0 or higher (available for Windows, Mac OS X, and Linux): https://cloud.r-project.org/
• R package VGAM: VGAM on CRAN
• MICC: MICC usage page. MICC download is the first link on the page.

Test installation by running the following line in R: library(MICC)

CID usage

java -Xmx10G -jar gem.jar CID --h

CID - Chromatin Interaction Discovery, version 1.0

Usage: java -Xmx32G -jar gem.jar CID --data INPUT-BEDPE-FILE --g CHROMOSOME-SIZE-FILE [--micc NUM] [--ex CHR] [--out FILE-NAME-PREFIX]
Options:
        --data INPUT-BEDPE-FILE
                the file path to the aligned paired-end bedpe or bedpe.gz file

        --g CHROMOSOME-SIZE-FILE
                the file path to the chromosome size file

        [--micc NUM]
                the minimum PET count of candidate interactions for the MICC input BEDPE file. Default NUM = 1. For large datasets, try NUM = 2, 3, or 4.

        [--ex CHR]
                the list of chromosomes to exclude. Default CHR = M. Multiple chromosomes can be separated by commas, e.g., CHR = M,Y
        
        [--out FILE-NAME-PREFIX]
                the output file name prefix
                
        [--h]
                Show this help message and exit

Example

• Install CID and MICC as described in the previous section
• Download the chromosome size file for genome hg19: hg19.chrom.sizes file
Chromosome size files for other genomes can be downloaded from the GEM software page or from the UCSC website.
• Download and unzip the example BEDPE file from a POLR2A ChIA-PET data set in K562 cells: BEDPE file
• Call CID: java -Xmx10G -jar gem.jar CID --data Ruan.K562.POLR2A.rep1.chr20.rmdup.bedpe.gz --g hg19.chrom.sizes --out test-run
• Note: if "out-of-memery" error occurs, increase the value in the -Xmx option for java.
• Call MICC on CID output BEDPE file by running the following lines in R:

  library(MICC)
  cid.bedpe <- read.table("test-run.bedpe", sep = "\t", header = FALSE)
  MICCoutput(cid.bedpe, "test-run-micc-out.txt")
  

• genomic interactions with significance measure (FDR) can be found in test-run-micc-out.txt

CID output

CID generates a BEDPE file containing identified genomic interactions (*.bedpe).

BEDPE is a header-less tab-delimited text file, following the BEDPE format, with the following columns:

1. genomic location of anchor A - chromosome
2. genomic location of anchor A - start coordinate
3. genomic location of anchor A - end coordinate
4. genomic location of anchor B - chromosome
5. genomic location of anchor B - start coordinate
6. genomic location of anchor B - end coordinate
7. PET count between two anchor regions
8. total PET count in anchor A (left anchor region)
9. total PET count in anchor B (right anchor region)

MICC output

MICC creates one output file (in the example above, test-run-micc-out.txt). It is a tab-delimited text file with the following columns:

1. genomic location of anchor A - chromosome
2. genomic location of anchor A - start coordinate
3. genomic location of anchor A - end coordinate
4. genomic location of anchor B - chromosome
5. genomic location of anchor B - start coordinate
6. genomic location of anchor B - end coordinate
7. PET count between two anchor regions
8. total PET count in anchor A (left anchor region)
9. total PET count in anchor B (right anchor region)
10. -log10(1 - posterior probability)
11. False Discovery Rate (FDR) of interaction

Citation

High resolution discovery of chromatin interactions
Yuchun Guo, Konstantin Krismer, Michael Closser, Hynek Wichterle, David K Gifford
bioRxiv 376194; doi: https://doi.org/10.1101/376194

Contact

Post your questions, problems, or suggestions on our GEM3 GitHub page by creating a "New Issue".