Predicting number of loci

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:

Predicting RAD loci for ddRAD

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?

solution

#note this may not work on a Mac, try on TACC if it doesn't
cd in_silico_digest/


cat test_reference.fasta | tr -d "\n" | grep -o -E "CATG.{100,300}AATT" | wc -l
cat test_reference.fasta | tr -d "\n" | grep -o -E "CATG.{100,350}AATT" | wc -l

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.

How would you use this to get an idea of how many loci you are likely to find.

Search NCBI for a reference genome
Download it
Search for your specific restriction search parameters

Solution:

solution

#download the reference from NCBI
wget ftp://ftp.ncbi.nlm.nih.gov/genomes/all/GCF/000/222/465/GCF_000222465.1_Adig_1.1/GCF_000222465.1_Adig_1.1_genomic.fna.gz
 
#decompress it
gunzip GCF_000222465.1_Adig_1.1_genomic.fna.gz

#check the command on some different max insert sizes
cat GCF_000222465.1_Adig_1.1_genomic.fna | tr -d "\n" | grep -o -E "GCATGC.{300,400}CCGG" | wc -l
cat GCF_000222465.1_Adig_1.1_genomic.fna | tr -d "\n" | grep -o -E "GCATGC.{300,500}CCGG" | wc -l
cat GCF_000222465.1_Adig_1.1_genomic.fna | tr -d "\n" | grep -o -E "GCATGC.{300,600}CCGG" | wc -l
 
#This can be done in a loop:


#set up the sizes you want to check
seq 400 100 1000 > maxInserts.txt
less maxInserts.txt


#build a header for your output
echo -e "insert_range\tNsites" > prediction_results.txt


#loop through your max insert sizes to get predictions for the number of sites
while read max
 do nSites=$(cat GCF_000222465.1_Adig_1.1_genomic.fna | tr -d "\n" | grep -o -E "GCATGC.{300,${max}}CCGG" | wc -l)
 echo -e "300-${max}\t${nSites}" >> prediction_results.txt
done < maxInserts.txt
 
#look at the results
cat prediction_results.txt
 
#These values are somewhat low
#What about a four-nucleotide restriction enzyme?
#for NlaIII
cat GCF_000222465.1_Adig_1.1_genomic.fna | tr -d "\n" | grep -o -E "CATG.{300,400}CCGG" | wc -l