RAD-Tags vs adaptor sequence vs restriction enzyme?

[David Wolfson]

Genetics noob here. Sorry for the stupid question, but what are the differences between RAD-Tags vs adaptor sequence vs restriction enzyme? I just ran process_radtags on my fastq.gz files which have already been demultiplexed (so no barcodes). I'm getting seemingly very high numbers of ambiguous RAD-Tags (most of the reads) and wondering if STACKS is correctly recognizing my restriction enzymes.

(stacks) wolfs064@system76-pc:~/first_round_seq_copy/umgc/2022-q4/221019_VH00601_82_AACCL5HM5/Reddy2_Project_001$ process_radtags -p . -o /home/wolfs064/stacks_output/first_plate/output_process_radtags/  --renz_1 pstI --renz_2 mspI -c -q -i gzfastq

No barcodes specified, files will not be demultiplexed.
Processing single-end data.
Using Phred+33 encoding for quality scores.
Found 104 input file(s).
Barcode type unspecified, assuming unbarcoded data.
Processing file 1 of 104 [20_S39_R1_001.fastq.gz]
  Processing RAD-Tags...1M...2M...3M...4M...5M...6M...
  6294438 total reads; -0 ambiguous barcodes; -5355730 ambiguous RAD-Tags; +0 recovered; -7222 low quality reads; 931486 retained reads.

[Catchen, Julian]

Radtags all begin with the restriction enzyme cutsite remnant (unless there is an inline barcode on the sequence prior to it, which you said there is not). The P2 adaptor can appear on the 3’ end of the sequence if you sequence through the insert length of the RADtag. The software is expecting the first five bases of each single-end read to be the pstI cutsite remenant, TGCAG. You should check if that is intact in your reads. Some ddRAD providers are using altered libraries where this is not the case, so you should check with who produced your libraries to understand what should be present.

 

It may be helpful for you to read one of the RADseq reviews that describe these details, like Andrews, et al. 2016, or Davey, et al. 2011.