I guess you need to provide more information whether you will use a collection for GWAS or you Will use a bi-parental population, if the later, which generation?  Have you tested any markers to test variation in your population? 

For a plant bi-parental population the "minimum" number would depend on a number of factors, including how precisely delimited and positioned you need your QTLs to be, how much phenotypic variance exists between the parental lines, and the algorithms used to calculate the QTLs.

Here's a paper that suggests populations of 200 for unbiased estimations of QTL position and effect:

https://www.researchgate.net/publication/44590811_Statistical_properties_of_QTL_linkage_mapping_in_biparental_genetic_populations

Article Statistical properties of QTL linkage mapping in biparental ...

Im using bi-parental population derived from the cross of Redgauntlet x Hapil (acronym RGxH) using "F1 cross- or more accurately a “two-way pseudo-testcross”". They were chosen as they showed a segregation in flowering time. 

I phenotype the population over 2 years. For 1st year i phenotyped 73 individuals and in the 2nd 63 individuals. I think i can get some QTL as the population showed transgressive segregation for the quality traits o interest, but would this work be accepted to be published !

See if this may help you https://www.researchgate.net/publication/268747417_Linkage_Mapping_and_QTL_Analysis

Chapter Linkage Mapping and QTL Analysis

Generally, a minimum of 100 individuals is considered for example by Euphytica, TAG, and others. Apart from this, with such limited number and using acceptable marker density you can only be confident about those major QTLs with highly significant LOD score. If you can increase the phenotypic power (number of individuals), you will find out how significant the change of your results may be. Some QTLs detected at LOD score close to threshold may not be detected and, on the other hand, new peaks could appear with above-threshold LOD.

As you already have the data and not planning for new experiments, I would suggest to run the analysis first and see what you get, then you can decide which journal to choose.

The link below will help you get an answer to your question

http://statgen.ncsu.edu/zeng/QTLPower-Presentation.pdf

The paper "Effect of population size on the estimation of QTL: a test using resistance

to barley stripe rust" by Vales et al. may also be helpful 

https://www.researchgate.net/publication/7584066_Effect_of_population_size_on_the_estimation_of_QTL_a_test_using_resistance_to_barley_stripe_rust

Article Effect of population size on the estimation of QTL: a test u...

Thank you so much for your replies. They are really helpful and i'll try to update you once i finished data analysis.

Many thanks ...

Hello Rashed 

My name is Reynaldo from Chile, Did you publish your result? just let me know

Hello Rashed

Iam trying to carry out a qtl analysis with a population size of 72 genotypes bi-parental population. Kindly share the outcome of your analysis if it would be possible for me to do the analysis with such population size.

Thanks

Hello Mercy,

Whatever your population size is, you can perform the marker trait association analysis to detect QTLs for your target traits. Of course if you have suitable marker density. But your population size would affect the reliability of your results. For example, in my research I anlized with a population size of 50 F2 plants and then increased into 100 plants and performed the analysis again. Some QTL peaks detected at the lower size disappeared when the population size was increased. And some others were further confirmed. On the other hand, some QTls were not detected with 50 F2 plants analyses but they were significant with 100 plants.

Some journals considers 100 indiciduals is a minimum size to consider the QTL analysis and the paper for publication. And this is widely accepted I guess.

Good luck with your analysis.

Thank you Mazel for your quick reply. I will consider this and go ahead with the analysis not for publication but maybe for my thesis.

Many Thanks

You are welcome

My question is that must the population to be used for QTL be from a cross on any line or single population

What is the minimum number of polymorphic SSR marker we can use to map a DH population of hexaploid wheat?