# How to decide if genes are linked in an unordered mapping problem?

From what I understand, if the tetrads resulting from a cross exhibit the same number of parental ditypes as there are nonparental ditypes, then the genes under consideration are said to be unlinked.

For example, say I cross two gametes containing $ab^+$ and $a^+b$, respectively, and get 170 tetrads exhibiting parental ditype (all offspring are $ab^+$ or $a^+b$), 165 exhibiting nonparental ditype (offspring are $a^+b^+$ or $ab$), and 30 exhibiting tetratype (offspring one of $a^+b^+$, $a^+b$, $ab^+$ or $ab$).

I'm not sure whether to classify these genes as linked or unlinked. I want to say yes, they are indeed linked, but 170 and 165 seem like close enough numbers to suggest they may be unlinked.

I would think the ratio of parental to nonparental ditype has to be considered close enough to 1, but how do we decide a threshold for that?

I'd use a chi-squared test. You calculate what you expect to see if your hypothesis is true, and you take what you do see, and the chi-squared result tells you how likely it is that your results are different enough from what's expected to indicate that your hypothesis is incorrect.

• Okay that makes sense to me. Is there a way to know the expected distribution for the tetrad genotypes? Feb 5, 2015 at 20:41