R package MeDEStrand (bedpe input available)
read1 read2 read1 read2 read1 read2 read1 read2 read1 read2 read1 read2 read1 read2
CHR STAR END CHR START END FRAGMENT ID MAPQ MAPQ STRAND STRAND CIGAR CIGAR FLAG FLAG TLEN TLEN NM_TAG NM_TAG
chr1 10028 10098 chr1 10152 10221 NB551056:129:H7... 0 0 + - 70M 24M1I45M 99 147 193 -193 3 2
chr1 10035 10105 chr1 10174 10230 NB551056:129:H7... 0 0 + - 70M 14S56M 99 147 195 -195 1 0
chr1 10175 10245 chr1 10257 10328 NB551056:129:H7... 9 9 + - 70M 28M1D42M 99 147 153 -153 5 4
...
- lean .bedpe should be pre-filtered from full .bedpe above by CIGAR and FLAG columns
- only retain fragments where read1 CIGAR has only matches (e.g. 70M, and not 68M2S)
- isPaired = TRUE ( FLAG 0x1 set )
- isProperPair = TRUE ( FLAG 0x2 set )
- hasUnmappedMate = FALSE ( FLAG 0x8 unset )
- isUnammpedQuery = FALSE ( FLAG 0x4 unset )
- isFirstMateRead = TRUE ( FLAG 0x40 set )
- isSecondMateRead = FALSE ( FLAG 0x80 unset )
- isSecondaryAlignment = FALSE ( FLAG 0x100 unset )
- ( all above are consistent with original MeDEStrand parameters )
- lean .bedpe should then be transformed into tab delimited document with columns shown below
- QWIDTH is not used in MeDESTrand, keeping it just for consistency
read1 read1 read1 FAKE read2 abs
CHR STRAND START QWIDTH START TLEN
chr1 + 10029 1 10153 193
chr1 + 10036 1 10175 195
chr1 + 10176 1 10258 153
...