I crossed two lines of rice with contrasting leaf, culm and panicle traits. The F1 plants were almost halfway between the parents except for some quantitative heterosis. The F2 plants (currently grown) are showing a wide range of combinations between the phenotypes of the two parents. I tried to classify them based on single or multiple phenotypes but failed to do so, they are messy!
My question is, how many generations should I grow to obtain true breeding lines with narrower range of phenotypes that are stably inherited?
Thank you for your suggestions.
The number of generations to reach uniformity in a trait from a F1 may vary by its complexity and the heterozigozity between parental genotypes, if you have previous knowledge of this you may estimate it. however as a general rule it takes between 5 to 7 generations of selfing to achieve pure lines (97-99% homocigous).
Starting from F1 at least 6 generation to obtain highly stable lines. Of course it depend also by the heterozigosity of founders
It takes minimum 5 to 6 generation to obtain doner character in recipient line.
Dear colleagues:
Thank you for your helpful contributions. Let me take my point a step forward.
Would it be correct if I start selection or exclusion of some phenotypes based on my final target starting from the F3?
The probability of a genotype being homozygous at all loci can be calculated using the following:
Probability = 1 - (0.5)^n, where n is the number of generations of selection beginning with an initial cross. As an example, the probability of a selection being pure is 0.50, 0.75, 0.875, 0.9375, 0.96875, after the 1st, 2nd, 3rd, 4th, and 5th generations respectively.
Or conversely, the number of generations required to achieve a defined degree of purity can be calculated using the following:
n = ln(1-desired purity)/(ln(1-0.5))
For example, the probability of achieving purity levels of 0.99999, 0.9999, 0.999, 0.99, 0.9, are 16.61, 13.29, 9.96, 6.64, and 3.32 generations, respectively, or rounding up, 17, 14, 7, and 4 generations respectively.
These equations assume random non-assortive mating, no linkage between different parts of each chromosomes, no epistasis, no differential reproductive fitness for different genotypes, and lack of immigration from genetic sources that differ in allelic frequencies. These equations also go out the window if dealing with cytoplasmic genes.
Lastly, if you are referring to specific traits, and not the entire set of traits possessed by a selection, then degree of purity can be assessed using markers. So theoretically, if you had say 400 markers for all traits you consider important, you can eliminate all non-desired types after the second generation. The problem with this scenario is the large majority of us or possibly all of us do not possess markers for all desirable traits in rice or other crops and excluding the obvious traits such as different types of disease tolerance/resistance, semi-dwarf genes, etc., we don't know the underlying genetic and biochemical bases, which requires the use of probability approximations for purity.
- Badges
- Science topic
- Similar topics
- Biological Science
- Genetics