Some boundaries are local maximum of insulation score profile instead of local minimum about crane-nature-2015

[blajoie]

Sorry for the delay - yes that does not sound correct. Boundaries should only be local minima, not maximia. Have you since solved your problem? If you can share your input/output I can try to take a look - though this code is now quite stale and is no longer maintained.

[yezhengSTAT]

Hello, I simply wrote an R code to extract local minimum and continue with other analysis without outputing the boundaries. The input file is attached if it is helpful.
rep1.Uni.chr1.ICE_matrix.is520001.ids320001.insulation.txt

`TADstrength <- function(TADdelta){
## input is two column TADdelta frame: delta and indicator
## return is the TAD boundary strength vector
i <- 1
end <- dim(TADdelta)[1]
max <- 0
min <- 1000000
extremeList <- NULL
strengthList <- NULL
binList <- NULL
while(i<=end){
if(!(is.na(TADdelta$deltaSquare[i])) & TADdelta$deltaSquare[i] == 1){ ## skip beginning NA 0 and peak
while(TADdelta$deltaSquare[i] == 1){
if(max < TADdelta$delta[i])
max <- TADdelta$delta[i]
i <- i + 1
if(i > end | is.na(TADdelta$deltaSquare[i]))
break

        }

        extremeList <- c(extremeList, max)
        max <- 0
        
        if(i > end | is.na(TADdelta$deltaSquare[i]))
            break
           
        binList <- c(binList, i-1)
        while(TADdelta$deltaSquare[i] == -1){
            if(min > TADdelta$delta[i])
                min <- TADdelta$delta[i]
            i <- i + 1
            if(i > end | is.na(TADdelta$deltaSquare[i]))
                break
        }
        extremeList <- c(extremeList, min)
        min <- 10000000
        if(i > end | is.na(TADdelta$deltaSquare[i]))
            break
           
    }else{
        i <- i + 1
    }

}

for(i in seq(1, length(extremeList), by=2)){
    strengthList <- c(strengthList, (extremeList[i] - extremeList[i+1]))
}
return(list(strengthList=strengthList, binList=binList))

}`