Simulating multiple loci (sequences) using the DNA marker

[Austin Koontz]

Hello all, 

I am new to the fastSimcoal software, and I am trying to simulate scenarios using the DNA marker type that has multiple loci (i.e. separate sequences) of a specific length (say 100 bp). However, when I specify the parameters for this (see attached .par file), what results is a single locus (sequence), with a variable number of alleles (in the attached example, ~2,400), that's 2,000 bp of length. 

I recognize that with the DNA marker, the num loci parameter specified in the final lines of the .par file corresponds to sequence length. But is there a way to generate multiple, separate sequences? Altering the number of chromosomes, chromosome blocks, etc., hasn't done the trick for me.

Thanks very much, for any help!

-Austin

[Austin Koontz]

As a follow-up to this question: I've noticed that in the Arlequin files being generated for the DNA markers, the LocusSeparator field (in the [Profile] section of the file) is set to "NONE". For other marker types (for instance, MSATs), this field is set to "WHITESPACE". So, is there a flag I can include with my fsc call, in order to change this field in the Arlequin output for DNA markers? I've attached examples of each of these outputs, as a demonstration. 

Thanks,

-Austin

[Austin Koontz]

As an update on this point: I have tested this behavior in the latest version of fastSimcoal27 (v09; executable fsc27093; October 2022), and have found it to persist (R code here). When running DNA simulations, DNA blocks are concatenated together.

The only way I've been able to address this problem is by using the strataG R package to run fastSimcoal27. When fastSimcoal simulations are ran through strataG, DNA blocks are separated (as demonstrated in the script linked above). I do not know why this works in strataG but not in fastSimcoal.