Hi,

bedtools usage question.

I know that when running intersectBed between a bam file and user-defined a bed file, the coordinates in the bam are 1-based, and the coordinates in the user-defined bed file, well I'm assuming they should also be 1-based.

The starts in a bed output of bedtools will be 0-based though.

What if I first converted the bam file to a bed file using bedtools's bamToBed and then ran intersectBed with my 1-based bed? Do I need to first add 1 to the starts in the bed output of bam2Bed?

I'll give an example: I have an RNA-seq bam that I want to extract sub-sequences of the mapped read, which intersect a certain interval.

I cannot find a single bedtools command for what I need so I'm doing it in two steps:

1. Converting the bam file to a bed file using bam2Bed with the -split argument, so that each aligned portion is in a different row in the output
2. Intersecting the bed file from step 1 with the bed interval.

So, here's a read from the bam file:

molecule/100    0   chr6    41058396    60  1S1096M100N295M865S *   0   0   GTTCCTTCCTTCCTCTCTTACTCTACCTTGGTTTCTCCACCAAACTTTCTCTCTCCCCTTTCCCACCCTCTTTCTCTTACTTTCCCCCTCCCTGTCTTCCTTCCTTTCTCTCTTTTCTTCTCCTTCTTTCTCCTTTTGTATCTCTTGCTTTCTCCTTTGAAACACTTCATATGTGAGCTTTGTGAAGTAAGGAAACACGATGTGCTAAGGGCACCAATGAATGCTCACAGAATTTTTAAATTATTGACTTTAGCTCACTGTTTTACAGATGAGAAAATGGAGTGCTTAAAAATGAATTGCTGATTTCCACATAACATGACCTGTATGATCCACAGGTAATCTATATATTTTCTACATTTGTTTTGGCTCTTGAACTCTGGCTGACTTAGTCCTCACTTTATGGCCCAGGTTGGCCTTGAACTCAGCAATCTTCCTGTCTTAGCCTTCTGAGTGCTGGGATTATAGGAGTTAAGACCGCATACCTGTATTCAATTCTTTTTGCTCCTTTCAGAATATTAATATCCTAAGTGTGTGATTATTTTCTTCTTGCTTTGCACTGTGTTGGCATTTAATGTCTCAAACAAGTAAGCCACATTTGTCCTTCCGGGAAGGCTTGTGTTCTTATTTCTTCTACTGATGTAGTCTTATGTTCCTTCTGTGTCCAGTAAAACATACATGTACCACCTGGGTTTACTGTTTGGCGGTATTACCAAATTTTTTCTGTTTGTCTCTCTCTGTCTGTGTCTCTTGTGTCACTCTGTCTCTCTTTCTCTTCATGTCTTTGTCTTACTCAATCTGTTTCTGTCTTTGTCTCTCTCTTTTTCTTGCCCTTTCTCTCTCCTCTCCTCTCCCTTCCCCCCCCCACGTGTGTGTGTGTGTGTGTGTGTGTGTGTGTGTGTGTGTTTGTGTGTGTGTGTGTATTTTTGAATCAAGCACCTTTCTGGCTGTGGTGATAGACAGACAAATCACATGTCATCTACAGTTGAATCTACTACCTGGTCTGATAGCTGCAGTAGGTAGGGCTTATTTGCCCTGCCTTGACCCAACTATGGGCTGTAGGCTCCTAAGCTGTGTGGCCTTCTGCCTCTTGGGAATAGGCCCTTTGGAGACGGCTGTTTTCCAGACTCCAAACTATCATGTCACACAGGTGGGAAATGAAGTGTCTTTCAATTGTAAGCAAACTCTGGGCCACGATACTATGTATTGGTACAAGCAAGACTCTAAGAAATTGCTGAAGATTATGTTTAGCTACAATAATAAGCAACTCATTGTAAACGAAACAGTTCCAAGGCGCTTCTCACCTCAGTCTTCAGATAAAGCTCATTTGAATCTTCGAATCAAGTCTGTAGAGCCGGAGGACTCTGCTGTGTATCTCTGTGCCAGCAGCTAAGAGTATAAATTCTGGAAATACGCTCTATTTTGGAGAAGGAAGCCGGCTCATTGTTGTAGGATTGCGGCTTTCCTATGCAAGCCACCACAGCTCTCTTACATCTCAGTGTAGGAGTGAATGTGGAACATCAGAGGATCTGAGAAATGTGACTCCACCCAAGGTCTCCTTGTTTGAGCCATCAAAAGCAGAGATTGCAAACAAACAAAAGGCTACCCTCGTGTGCTTGGCCAGGGGCTTCTTCCCTGACCACGTGGAGCTGAGCTGGTGGGTGAATGGCAAGGAGGTCCACAGTGGGGTCAGCACGGACCCTCAGGCCTACAAGGAGAGCAATTATAGCTACTGCCTGAGCAGCCGCCTGAGGGTCTCTGCTACCTTCTGGCACAATCCTCGCAACCACTTCCGCTGCCAAGTGCAGTTCCATGGGCTTTCAGAGGAGGACAAGTGGCCAGAGGGCTCACCCAAACCTGTCACACAGAACATCAGTGCAGAGGCCTGGGGCCGAGCAGACTGTGGGATTACCTCAGCATCCTATCAACAAGGGGTCTTGTCTGCCACCATCCTCTATGAGATCCTGCTAGGGAAAGCCACCCTGTATGCTGTGCTTGTCAGTACACTGGTGGTGATGGCTATGGTCAAAAGAAAGAATTCATGAAGCCAGATGTGAAGATGAATACAAGAGCTGACAACACATTGTGTTAATACAGATTTTCTTCTCCCAGAACTTCTGAAGAGCTATTCTCATTTGTCTGTGCATCCCAAATTCTGCCTACTAGTCACGCATAGGTGCATTTGTATGTCTGAAATTCTTGTGACCTGGAAAATGCCTACACTTACAATCAAACCAATAAACATGTTCTAGGACGGCCG   *   NM:i:0  ms:i:1391   AS:i:1359   nn:i:0  ts:A:+  tp:A:P  cm:i:439    s1:i:1343   s2:i:0  de:f:0  SA:Z:chr6,41534323,+,1399S857M4700D1S,60,7; rl:i:204

The corresponding bed rows are:

chr6    41534322    41534372    molecule/100    60  +
chr6    41537511    41537583    molecule/100    60  +
chr6    41538217    41538605    molecule/100    60  +
chr6    41539225    41539335    molecule/100    60  +
chr6    41539644    41539879    molecule/100    60  +

So since bamTobed is outputting the bed in 0-based coordinates, the actual starts are the above + 1:

chr6    41534323    41534372    molecule/100    60  +
chr6    41537512    41537583    molecule/100    60  +
chr6    41538218    41538605    molecule/100    60  +
chr6    41539226    41539335    molecule/100    60  +
chr6    41539645    41539879    molecule/100    60  +

When I intersect these bed rows (the 0-based output) with my bed interval:

chr6    41538218    41539881    g1

I'm getting:

chr6    41538218    41538605    molecule/100    60  +   chr6    41538218    41539881    g1  100 +
chr6    41539225    41539335    molecule/100    60  +   chr6    41538218    41539881    g1  100 +
chr6    41539644    41539879    molecule/100    60  +   chr6    41538218    41539881    g1  100 +

So in the intersection output are the starts (second column) 0-based or 1-based?