I found methyAnalysis is a pretty nice tool after testing several packages for the same purpose.  But recently I found an issue. I used default setting to identify significant DMRs,  so the maximum gap allowed between two nearby probes should be no more than 2000. But why the resultant DMRs including neighbouring probes very far apart (even over 1Mp) ?  Thanks!

After looking into the code, I change two lines in the part of "check the gaps of sparse probe" and think it works. 

    ## check the gaps of sparse probe
    gaps.i <- which(c(FALSE, diff(allPosition.i) > maxGap) & status.i == 1)
    # the gap should be 1 in both sides
    gaps.i <- gaps.i[status.i[gaps.i - 1] == 1]  #### revised
    if (length(gaps.i) > 0) {
      startInd.i <- sort(c(startInd.i, gaps.i))
      endInd.i <- sort(c(endInd.i, gaps.i-1)) ### revised
    }

Now I find another problem of this program; there are some output DMRs with changes in opposite direction.  Is there anyone see the same problem?