In large scale breeding plant projects, first a list of desired quantitative and qualitative traits
are identified. Qualitative traits are things like taste, aroma, to some extent growth habit etc, Quantitative traits are those that can be measured, such as biomass, thc levels, flowering time etc.
Then once you have these attributes you can begin to look for your target plants; once lines are loosely identified; you need to understand the combining abilities ( and direction) of the crosses you are going to make if you hope to make the cross predictable.
There is two types of combining abilities:
1.General Combining Ability (GCA).
General combining ability (GCA) is directly related to the breeding value of a parent and is associated with additive genetic effects.
The GCA of a line is the difference of its average performance in hybrids relative to the grand mean of all line crosses, The expected mean of a cross is the sum of the maternal and paternal general combining abilities, GCA§ + GCA(M).
- Specific combining ability (SCA)
Specific combining ability (SCA) is the relative performance of a cross that is associated with non-additive gene action, predominantly contributed by dominance, epistasis, or genotype × environment interaction effects
In large breeding programs a full diallel design is often done involving dozens of lines and hundreds of plants, obviously that’s outside the abilities of the vast bulk of us, so there are a couple or approaches that can be taken:
The topcross design, in which a particular line (the testor) serves as the common male (pollen) parent and the GCAs of female plants are estimated by the performance of their seed. For this, inbred stable lines are needed otherwise it’s impossible to work out where a trait might have come from and those traits may not be reflected or fixed further down the line, besides it renders any objective analysis useless if random traits are popping up without any predictable continuity into next crosses.
The other approach is a polycross design wherein females are allowed to be randomly fertilized from among all the other lines. Again, the elite lines are chosen by those with the best-performing offspring.
From here you can gauge the combining abilities of the targets in the cross and work out early if they are traits that are going to be reasonably straight forward to fix, or very difficult etc. there’s a whole lot of statistical analysis tools that be used for working out the most efficient crossing that delivers both the desired traits and maximum hererosis/hybrid vigour etc.
Also the direction of a cross is an important factor in working out the nuts and bolts of how and when you are going to cross and which line is the male vs the female, they may be completely different.