I'm a horse pedigree nerd who likes to find common matings in the pedigrees of successful cutting horses.
I've started to use formulas in spreadsheets for this but it's becoming quite tedious to come up with a formula for each possible combination in a 5 generation pedigree.
I've been wondering whether I can use R to help me find breeding patterns (matings) that may repeat in a sample of successful individuals (I suspect there are several of such patterns). The goal is to be able to say that 'this pattern is present in winners of $X million' (earnings are included in the dataset).
The data (ancestors) for each individual are laid out in rows as follows:
Column 1: Horse's name (current performer)
Column 2: Money earned
Column 3: Generation 1 Top (name of sire)
Column 4: Generation 1 Bottom (name of dam)
Column 5: Generation 2 Top (name of paternal grandsire)
Column 6: Generation 2 Top (name of paternal granddam)
Column 7: Generation 2 Bottom (name of maternal grandsire)
Column 8: Generation 2 Bottom (name of maternal granddam)
And so on until the 5th generation.
Each following generation has the double of horses as the previous one.
I’m not interested in finding out the great producing individuals over the last few decades (everyone already knows who they are). I’m more interested about the crosses that tend to produce winners. And by crosses I don’t mean the sire and the dam, but the bloodlines a bit further back in the pedigree.
For example, there are cases in which two successful half brothers are out of different mares that are very closely related but not in an obvious way. They could share some ancestors in the 3rd or 4th generation that are placed similarly but not equally. There could be two individuals that are 3/4 siblings even though they are by and out of different individuals.
I hope this makes sense to a non-horsey person?
From my initial research into R seems like selecting cases using multiple selectors may do the job?
Or is there a base function in R or a package that could work better?
Just trying to get any useful intel before I go down a rabbit hole that takes me nowhere.