View Single Post
Old 02-13-2012, 11:27 PM   #2
Simon Andrews
Location: Babraham Inst, Cambridge, UK

Join Date: May 2009
Posts: 871

In a normal genomic paired end library there should be no preference for whether the first read of the pair maps to the original top or bottom strands. The only constraint should be that the second read should map to the opposite strand to the first and that the 3' ends of the two reads should point towards each other - usually separated by the approximate insert size of the library. If you are viewing these mapped pairs of reads in a genome browser then the read on the left (most 5' on the top strand) could have come from either the first or second read of the actual sequencing run.

There are library preparation methods (mostly RNA-Seq or ChIP-Seq) where a single strand is selected, so the coverage of a region will be strand biased, but this won't be a global effect (ie everything on the same genomic strand), but will be localised to a specific transcript or binding site.
simonandrews is offline   Reply With Quote