Parental or

The easiest way to get pure-breeding lines is to get them from a stock center. However, this only works for a few organisms. In this case, some other geneticist has done the work for you.

Observation of Populations: Most organisms of interest to geneticists have been in laboratory or field culture for many generations. These organisms have been maintained by mass mating (e.g., Drosophila) or by systematic breeding systems (mice). Careful scientists observe the phenotypes (for example,) eye color) of many progeny in each generation. If all progeny have the same phenotype as their parents (breed true), then the scientists suspect that the population is homozygous for that specific phenotype.

This is not 100% accurate and does not work well with organisms that have very few progeny.

Selfing: Some organisms can donate both eggs and sperm (selfing). Allowing organisms to self over about 10 generations guarantees about 99% homozygosity at all loci. In practice, if the offspring of selfed individuals never show an alternative phenotype, it is safe to assume that the original individual was homozygous for the trait in question. This works quite well for C. elegans.

Brother-sister mating: Most individuals cannot reproduce by selfing. However, you can get homozygous lines by mating close relatives. Starting with a single ancestral breeding pair, a brother and sister are chosen each generation. Inbred strains take 20 generations of brother-sister matings to achieve 99% homozygosity at all loci. However, with regard to a single locus with a defined phenotype, if the experimenter does not see an alternative phenotype for that trait over the first 5 to 10 generations, it is very likely that the original brother and sister pair was pure breeding and the line can be considered homozygous for that locus.

Alternatively, inbred lines can be established by mating offspring back to a parent in a regular pattern.

A new phenotype occurs in a population. This might be due to a spontaneous mutation or it might be due to an induced mutation. At this point, nothing is known about the inheritance of the trait.

The first step is to preserve the potential allele by breeding. The type of breeding will depend upon the number of organisms with the new phenotype, the breeding system, and the gender of the organism with the new phenotype.

Observation 1: In a wild-type population of lab organisms a new phenotype shows up in a number of progeny. You have both males and females exhibiting this trait. In this case, it is easiest to mate the males and females with the trait to see if you can establish a pure-breeding line. If the trait is recessive, all of the progeny will have the recessive phenotype. If the trait is dominant, all of the individuals should be heterozygous and 3/4 of the progeny will show the trait.

At the same time, it is desirable to mate as many potential mutant organisms to wild-type organisms (preferably from a stock not known to be segregating the new mutation) to preserve the gene. If the trait is recessive, none of the progeny should show the new phenotype. If the trait is dominant, 1/2 of the progeny will show the new phenotype.

Example:

As a graduate student at the University of Illinois I noticed a number of orange-eyed blow flies (Phormia regina) in one of my stocks. Phormia all have red eyes. I mated several pairs of orange-eyed flies in an attempt to get a pure breeding line. At the same time, I mated several orange-eyed males to many red-eyed females in order to preserve the mutation.

The cross of orange x orange produced only orange offspring. The cross of orange male to wildtype red female (or wildtype females to orange male) produced only red-eyed progeny. The results of the cross were consistent with orange eyes being a recessive trait.

In this case, the orange x orange cross produced several hundred progeny. After one generation, I had a pure-breeding line of orange-eyed flies. However, the progeny of this stock continued to be observed each generation to be sure that only orange-eyed flies were produced.

Observation 2: A single individual with a unique phenotype shows up in the population and it is a male.

Mutant males are great since, in general, they can be mated to multiple wildtype females. If the trait is recessive, none of the progeny should show the new phenotype. If the trait is dominant, 1/2 of the progeny will show the new phenotype. (slightly different rules may apply if the trait is sex-linked).

Observation 3: A single individual with a unique phenotype shows up in the population and it is a female.

The females should be mated to several males. The interpretation of the outcome will depend on the fecundity of the female. The two most likely cases are that the trait is recessive and the female is (aa) or that the trait is dominant and the female is (Aa). Again, if the trait is recessive, none of the progeny should show the new phenotype. If the trait is dominant, 1/2 of the progeny should show the new phenotype. (Slightly different rules may apply if the trait is sex-linked). However, a certain number of progeny must be produced to interpret these data.

Female is heterozygous and 1/2 of the progeny should have the new phenotype
Progeny Observed	Wild-type Progeny Only	Probability
1	1	.5
2	2	.25
3	3	.125
4	4	.0625
5	5	.0313
6	6	.0156
7	7	.0078
8	8	.0039
9	9	.0020
10	10	.0010

We need at least 7 progeny to have better than a 99% of detecting a dominant allele. Obviously 10 or more progeny would be better. On the other, hand even a single offspring is informative if the recessive phenotype shows up.