Output format of nucleatide simulation

[Zuxi Cui]

Hi Ben,

I have a suggestion for further updates on SLiM. I've been using SLiM for simulating genome-wide data for a while and realized the output format of ".vcf" is not an efficient one. The random mating system does not take long to process but writing data as a VCF takes a lot of time. For example, it took me less than an hour to process the mating but more than 300 hours to write 13,000 samples of a single chromosome (chr15) as the VCF is not compressed at all (2376GB). PLINK2 format (binary pedigree) is much more efficient than VCF for us researching whole-genome data. After transforming, for the same dataset described above, the PLINK2 files (pgen+pvar+psam) are less than 40GB. Besides, another popular genome-wide simulator Hapgen2 also uses a more efficient file format (haps+legend). The haplotype format is less efficient than PLINK2 since it was developed around 2011 but is still ~5x more efficient than VCF.

Can you consider adding other output formats for genome-wide data in the future? It will be very helpful and timing-saving for us users.

Thanks,

Terry

[Peter Ralph]

Hi, Zuxi! Good suggestions. You might look at the tree sequence output
option, which is very efficient and can be easily output to genotype
matrices, etcetera.

[Zuxi Cui]

Thanks for the tip. Is there a chapter in the manual that I can follow? I prefer some sample codes to start with.

[Peter Ralph]

Well, Chapter 17 of the SLiM manual talks about tree sequence
recording in SLiM; then you might look at these tutorial for how to
efficiently analyze the results:
https://tskit.dev/tutorials/getting_started.html#saving-and-exporting-data

If you don't find it and have a suggestion for what would be a good
simple example of the sort of thing you want to do, let me know!
* peter