Cannot detect my simulated bottleneck

[JaneeshBansal]

Hi,

I am having trouble identifying a bottleneck that I simulated.

I simulated data using msprime with the following values ...

• Parameter values: 30Mbp genome, 50 chr, recomb rate = 5e-9, mut rate = 2.5e-8, generations = 800,000
• Bottleneck model: [0 : 200,000 gens] Ne = 30,000 , [200,000 : 800,000] Ne = 60,000

I first sampled a single diploid individual, however, the plot did not show a bottleneck. From the paper, I understood that 2 lineages are part of the PSMC and the n-2 are part of the SFS. But if I sampled a single individual then only the PSMC part of SMC++ is being ran?

I also ran my simulation through MSMC2 running PSMC' and the plot did show a bottleneck shape.

Additionally, I sampled 5 and 50 diploid individuals but the characteristic bottleneck shape was not appearing in SMC++.

Could I please receive some advice on this?

Thank you Janeesh

[terhorst]

It looks as SMC++ is fitting a single constant piece over the whole region 1e5~7e5 generations. Try specifying the timepoints manually using the --timepoints option as described in the README.