Re: Sample size and marker number for structure...

[Apurba Barman]

Hello,

I was curious to know if there is any way to tell how may samples and/or markers should provide signal of population structure (if there is any). Of course, it might vary according to the type of marker used.

Any input is appreciated.

Apurba

[Vikram Chhatre]

Hi Apurba,

I don't think there are any specific guidelines as far as number of
markers and individuals that can give maximum information about
population structure. Broadly speaking, more is better. However, with
genome-wide data, you may get similar answers when using a small
proportion of markers compared with really large number of markers,
both spanning the genome. Of course, numer of markers and individuals
is proportional to the amount of computational time, although that
process could be parallelized (not in a strict sense of the word) by
running multiple instances of Structure if you have a cluster
available.

You could try exploratory analyses using different numbers of markers
and individuals in your dataset to see if you get similar answers in
both cases.

All the best
Vikram

> --
> You received this message because you are subscribed to the Google Groups
> "structure-software" group.
> To post to this group, send email to structure...@googlegroups.com.
> To unsubscribe from this group, send email to
> structure-softw...@googlegroups.com.
> For more options, visit this group at
> http://groups.google.com/group/structure-software?hl=en.

[Adii_ (Institute of Genetics and Animal Breeding)]

I guess it also depend on Your population structure. If there are big differences You will discover this population with smaller amount of data. 

By the way, it could be interesting to try to estimate minimum sample size according to different structure models :)