FreeHiC simulator

This function performs FreeHiC based on contact matrix only.

FreeHiC(
  contacts,
  seqDepth = NULL,
  countScale = 1,
  noiseRate = 0,
  neighborZeroRate = 0,
  resolution = 50000L
)

Arguments

contacts	A list of contacts matrix. The list element should named by chromosome pair. matrixA matrix with columns (x, y, counts). data.frameA data.frame with columns (chr1, x, y, chr2, counts). listA list of matrix, where the matrix has columns (x, y, counts). where x stands for the first chromosome location, y stands for the second chromosome location and counts is the interaction counts.
seqDepth	The desired sequence depth.
countScale	The scale of counts. A number larger than 0. If both seqDepth and countScale are provided. Choose the larger one.
noiseRate	The noise rate for contact matrix. 0 - 1 scale
neighborZeroRate	The rate for neighborhood noise rate. 0 - 1 scale
resolution	The resolution used in the contacts matrix. A positive number.

Value

A list or matrix or data.frame with the same format as contacts

References

Zheng, Ye, Keles, Sunduz FreeHi-C simulates high-fidelity Hi-C data for benchmarking and data augmentation. Nature Methods 17, 37–40 (2020). doi

Examples

library(FreeHiCLite)

N <- 2000
maxBinX <- maxBinY <- 2000000L
maxCounts <- 10
binX <- sample(1:maxBinX, N, replace=TRUE)
binY <- sample(1:maxBinY, N, replace=TRUE)
counts <- sample(1:10, N, replace=TRUE)

seqDepth <- 20000
countScale <- 0
noiseRate <- 0.1
neighborZeroRate <- 0
resolution <- 5000

## Matrix version
## matrix layout as x, y, counts
contacts <- matrix(0, N, 3)
contacts[,1] <- binX
contacts[,2] <- binY
contacts[,3] <- counts

res <- FreeHiC(contacts, seqDepth, countScale, noiseRate, neighborZeroRate, resolution)
head(res)
#>            x    y counts
#> [1,] 1902912  988     16
#> [2,] 1993687 1073     12
#> [3,] 1047284 1341      4
#> [4,]  922396 2227     10
#> [5,] 1647076 3340     13
#> [6,] 1673148 4468     11

## List version
N2 = 3000
binX2 <- sample(1:maxBinX, N2, replace=TRUE)
binY2 <- sample(1:maxBinY, N2, replace=TRUE)
counts2 <- sample(1:10, N2, replace=TRUE)

contacts2 <- matrix(0, N2, 3)
contacts2[,1] <- binX2
contacts2[,2] <- binY2
contacts2[,3] <- counts2

contactsMap <- list('1_1' = contacts, '2_2' = contacts2)
res <- FreeHiC(contactsMap, seqDepth, countScale, noiseRate, neighborZeroRate, resolution)
str(res)
#> List of 2
#>  $ 2_2: int [1:2715, 1:3] 1438739 1515769 226756 1285103 909010 27773 1219428 1046613 1530112 1023819 ...
#>   ..- attr(*, "dimnames")=List of 2
#>   .. ..$ : NULL
#>   .. ..$ : chr [1:3] "x" "y" "counts"
#>  $ 1_1: int [1:1829, 1:3] 1902912 1993687 1047284 922396 1647076 1673148 934434 1914771 1043449 46432 ...
#>   ..- attr(*, "dimnames")=List of 2
#>   .. ..$ : NULL
#>   .. ..$ : chr [1:3] "x" "y" "counts"

## Dataframe version

chr1 <- c(rep('1', N), rep('2', N2))
chr2 <- c(rep('1', N), rep('2', N2))

contactsAll <- rbind(contacts, contacts2)

contactsDf <- data.frame(chr1 = chr1, x = contactsAll[,1],
chr2 = chr2, y = contactsAll[,2], counts = contactsAll[,3])

res <- FreeHiC(contactsDf, seqDepth, countScale, noiseRate, neighborZeroRate, resolution)
head(res)
#>   chr1       x chr2    y counts
#> 1    2 1438739    2    8      3
#> 2    2 1515769    2 2134      5
#> 3    2  226756    2 2376      2
#> 4    2 1285103    2 3500      9
#> 5    2  909010    2 3950      8
#> 6    2   27773    2 4715      1