Epistasis + Chi-squared

What is epistasis?

Interaction of (products of) non-linked genes where one masks / suppresses the expression of the other.

Describe when chi-squared can be used.

• When determining if observed results are significantly different from expected results (frequencies)

  • Eg. comparing the goodness of fit of observed phenotypic ratios with expected ratios

• Data is categorical (can be divided into groups eg. phenotypes)

Suggest why in genetic crosses, the observed phenotypic ratios obtained in the offspring are often not the same as the expected ratios.

• Fusion / fertilisation of gametes is random

• Autosomal linkage / epistasis / sex-linkage

• Small sample size —> not representative of the whole population

• Some genotypes may be lethal (cause death)

Describe how chi-squared value can be calculated

x² = (O-E)² / E

O = frequencies observed

E = frequencies expected (multiply total n with each expected ratio as a fraction)

How a chi² value can be analysed

1. Number of degree of freedom = number of categories - 1 (eg. 4 phenotypes = 3 degrees of freedom)

2. Determine critical value at p = 0.05 from a table

3. If x² value is (greater / less) than critical value

   • Difference (is / is not) significant so reject / accept null hypothesis

   • So there is (less / more) than 5% probability that difference is due to chance.