Home Contact
no

Introduction

scEChIA (Single Cell Epigenome Chromatin Interaction Analysis) is a computational method that can predict chromatin interactions among distal sites with high accuracy. scEChIA can also make UCSC track on predicted interactions that may be useful for genomics studies based on chromatin interactions using the UCSC Genome Browser.

Installation

Download Package

or use the link http://reggen.iiitd.edu.in:1207/scEChIA/scEChIA_Package/scEChIA_0.1.0.tar.gz

  • version.string R version 3.6.0
  • These libraries need to be installed before the package is installed : varbvs, pracma, glassoFast, nor1mix, roxygen2.
  • install.packages("scEChIA_0.1.0.tar.gz", repos=NULL, type="source")


    • Library required for scEChIA package:

  • library(glassoFast)
  • library(nor1mix)
  • library(roxygen2)
  • library(scEChIA)
  • library(varbvs)
  • library(pracma)


    • Chromatin interaction prediction using prior rho information:

    Function_name: Interaction_Prediction_1

  • Interaction_Prediction_1 can predict interaction with the help of two different cell types' rho value. First we will pass HiC data from two different celltypes and take average of both HiC data to define as final rho matrix for predicting interaction from single cell data.

    • How to run:

    setwd(system.file("data", package="scEChIA")) #Keep your input data format like example data
    genomic_region1 = read.table('K562_HiC_chr17.txt')
    genomic_region1[is.na(genomic_region1)] = 0
    genomic_region2 = read.table('IMR90_HiC_chr17.txt')
    genomic_region2[is.na(genomic_region2)] = 0
    gap = 25000 #Bin size 25kb (User can change it as per his choice)
    patternf = 17 #Chromosome number (User can change it according to chromosome number)
    chrNo = patternf
    data = read.table('GM12878_SC_chr17.txt')
    chrinfo = data[, 1:3] #Chromosome location (chr, start, end)
    chrNo = patternf
    startCell = 96 #start sample column
    endCell = 575 #end sample column
    chromSize = 83257441 #Size of chromosome 17 (Change it according to chromosome number. Follow heading Chromosome Size of hg19)
    rhomatrix = rhomatAvg(genomic_region1, genomic_region2, gap, patternf, data, chrinfo)
    predicted_interaction = Interaction_Prediction_1(chrinfo, data, rhomatrix, chrNo, startCell, endCell, chromSize)

    Chromatin interaction prediction using constant rho:

    Function_name: Interaction_Prediction_2

  • Interaction_Prediction_2 can predict interaction with the help of constant rho value.

    • How to run:

    gap = 25000 #Bin size 25kb (User can change it as per his choice)
    patternf = 17 #Chromosome number (User can change it according to chromosome number)
    chrNo = patternf
    data = read.table('GM12878_SC_chr17.txt')
    chrinfo = data[, 1:3] #Chromosome location (chr, start, end)
    chrNo = patternf
    startCell = 96 #start sample
    endCell = 575 #end sample
    chromSize = 83257441 #Size of chromosome 17 (Change it according to chromosome number. Follow heading Chromosome Size of hg19)
    rhoConstant = 0.01 #user can change the value of rhoConstant
    predicted_interaction = Interaction_Prediction_2(chrinfo, data, rhoConstant, chrNo, startCell, endCell, chromSize)

    Visualizing chromatin interactions using UCSC track

    Function_name: ucscTrack

  • Function ucscTrack can create track file for UCSC Genome Browser on predicted interactions. Using addscore value it will provide perfect interaction score in decimal numbers.

    • How to Run :

    predInteraction = predicted_interaction
    addscore = 500 #User can increase or decrease addscore value as per his choice
    ucscTrack_file = ucscTrack(predInteraction, addscore)

    Note :

    After generating the UCSC track file, the bottom line has to be pasted on the header to be visualized on the UCSC Genome Browser. track type=interact name="User and celltype" description="User and celltype" interactDirectional=true maxHeightPixels=200:100:50 visibility=full

    #chrom chromStart chromEnd name score value exp color sourceChrom sourceStart sourceEnd sourceName sourceStrand targetChrom targetStart targetEnd targetName targetStrand


    Visualization figure of microglia using UCSC Genome Browser (https://genome.ucsc.edu/) :

    no

    How to validate predicted chromatin interactions? :

    1. Predict the chromatin interaction using scATAC-seq data from scEChIA. 2. Find the related available chromatin interaction from HiC/chromatin-based data and make it paired for every interaction. 3. Use freely available resource - paired genomic loci tool ( pgltools - https://github.com/billgreenwald/pgltools) to intersect predicted chromatin interactions with existing chromatin interactions. 4. Check the fraction/number of overlap.


    Chromosome Size of hg19

    chr1: 248956422, chr2: 242193529, chr3: 198295559, chr4: 190214555, chr5: 181538259, chr6: 170805979, chr7: 159345973, chr8: 145138636, chr9: 138394717, chr10: 133797422, chr11: 135086622, chr12: 133275309, chr13: 114364328, chr14: 107043718, chr15: 101991189, chr16: 90338345, chr17: 83257441, chr18: 80373285, chr19: 58617616, chr20: 64444167, chr21: 46709983, chr22: 50818468, chrX: 156040895, chrY: 57227415

    Supplementary Files