An important question in planning a RAD-seq experiment is how many loci to expect from your organism's genome
If you have a reference genome for your organism, (or a closely related one) it is easy to get a prediction.
Although you should take into consideration how much of the actual genome is represented in the reference.
From John Stanton-Geddes and seqAnswers:
cat pbar_scaffolds_v03.fasta | tr -d "\n" | grep -o -E "CATG.{264,336}AATT" | wc -l
Where pbar_scaffolds_v03.fasta is your reference genome
In "CATG.{264,336}AATT" replace CATG with the forward restriction enzyme recognition sequence and AATT with the reverse restriction enzyme recognition sequence.
This example would work if your foward restriction enzyme is NlaIII, your reverse restriction enzyme is MluCI, and your insert size is between 264 and 336 base pairs.
Does this really work?
This fake reference has 4 RAD fragments, 3 with 100-300 bp insert sizes and 1 more with a 350 bp insert size for NlaIII and MluCI
Does the method return correct answer?
Exercise:
Imagine you wanted to do ddRAD on the coral Acropora hyacinthus
There is no reference genome for this coral, but there is one for a closely related species Acropora digitifera.
Your lab has the restriction enzymes Sphl and MspI on hand and you want to use those as your forward and reverse cutter respectively.
You are sequencing three sample populations and based on power analyses you have decided you'll need sample population sizes of n=20 (N=60 total samples).
Finally, you have decided to do 150 bp paired end sequencing on the Illumina Hiseq 4000 and have exactly $3,239 dollar to dedicate to sequencing.
Assuming you will get 240 million reads from a single lane of an Illumina Hiseq 4000 run, an a minimum insert size of 300, what should you pick your maximum insert size as?
- Search NCBI for a reference genome
- Download it
- Search for your specific restriction search parameters
Solution: