Why do reverse radtags have different start points in radtag sequencing?

Question

Why is it that when doing paired end Radseq that R1 and R2 have different start points? For instance, while doing Radseq you do a restriction digestion of your DNA using enzyme X- shouldn't the sequencing reaction take place from either end of where the enzyme cut? If it doesn't- but happens randomly where do your multiplex sequences go?

For instance- if you are sequencing the sequence on an illumina machine with the "<>" representing the cut site-

Where does the sequencing reaction start from in each direction?

Where does the multiplexing sequence go? - does it go on each?

5-GGCTAGC<CCTGCA*GG>GCGCTT-3 3-CCGATCG<GG*ACGTCC>CGCGAA-5

Answer 1

Assume that after digestion you're left with the following sequence:

5' AGCTAGCTAGCTAGCTGACGTCGTAC 3'
3' GTACGACGTCAGCTAGCTAGCTAGCT 5'

Then one read would start with AGCTAG and the other would start with TCGATC.

The barcodes (what are used for multiplexing) are outside of this and part of the illumina adapters. They're sequences in completely different reads. This is nicely illustrated below:

Index 1 and 2 don't actually need to be present. Even if they are, whether the sequencer actually sequences them is a user setting.

For those curious, the order of the reads as they're sequenced are:

1. read #1
2. Index 1 (optional)
3. Index 2 (optional)
4. read #2 (optional)

If you ever get more than 2 reads from a sequencing provider, then it's likely that they included the index sequence (I have no clue why, but some sequencing centers like to do this...even though it confuses the heck out of their users).