Question about how fragments are sequenced

[blakeoft]

I think I have a pretty good understanding of library preparation for next gen sequencing, but I have a couple quick questions. These are roughly the steps that I understand:

1. DNA randomly fragmented
2. Ends of fragments are repaired and adapters ligated to the end
3. Fragments denatured (this doesn’t seem to be mentioned in any of the diagrams of videos I’ve seen, but I guess it's implied)
4. Single strands are hybridized randomly to a flow cell on which several adapters are attached
5. Nucleotides and polymerase construct reverse strands
6. Fragment denatured so that only the reverse strand remains and is attached to the cell
7. Bridge amplification occurs several times until clusters are large enough (or maybe it just happens a certain number of times -- I’m not too worried about this)
8. All reverse strands are washed away leaving only the strands that match the strands attached in step 4
9. Tagged nucleotides bind to the strand and release some kind of light signal that is detected by a camera

My question has to do with step 9. Let’s say the sequencer picks up a T, and we assume that it’s the right call. Wouldn’t this T be part of the reverse strand? If so, do aligners like BWA take this into account and try to align reverse complements of reads as well? Also, how are the single strands chosen in step 4? Is that just random?

Thanks.

[SNPsaurus]

Let's say you have a genomic fragment CAT (double stranded CAT/GTA) and attach simple adapters to it:

TTTCATAAA
AAAGTATTT

Now it hybridizes to the AAA flowcell oligos. Note that both the top and bottom strand can hybridize to AAA with 5' TTT adapters, and one will sequence CAT and other sequence as GTA. So there is no expectation of directionality for which strand will be sequenced compared to a reference sequence. Aligners take this into account, and will report the alignment as either using the forward or reverse direction.