Part I

1. There are two different techniques used to obtain a pooled estimate of the recombination fraction (p). A weighted estimate of p using the information content is the first method and Fisher's scoring method is the second technique. The advantage to Fisher's scoring method is that a test for homogeneity of the sources of data can be evaluated.
   
   
2. Whe we have separate estimates of the recombination fraction (p) using different types of families, we can combine the separate estimates into an overall estimate using a weighted p value. The weights for each estimate of p are the information content of
   each family where the combined p = Sp_I
   each family where the combined p =  S^I
   We can then determine the weighted standard error of the combined p estimate using V_p = SI_i,
   se(p) = the square root of V_p, which equals the square root of SI_i
   
   
3. Fisher's scoring method is used to obtain a test of significance of the deviations from independent assortment for each type of family evaluated and for the pooled estimate of p. This method also provides a test of the homogeneity of the various sources of data to be combined and a pooled estimate of p.
   
   
4. The score using Fisher's scoring method is obtained from the unreduced maximum likelihood equations using the last estimate of p in the formula to determine the score. A Chi-square test can be conducted to evaluate the fit of the p estimate to the observed data.
   
   
5. For testcross data, use direct calculation of precent recombination in the population which is n_r/n = p. This formula to estimate p from testcross data is actually the maximum likelihood estimate of p.
   
   
6. For F₂ coupling or repulsion data alone, use the product method to estimate p.
   
   
7. For F₃ data from doubly dominant F₂ phenotypes, the maximum likelihood method must be used to estimate p.
   
   
8. For combined data from F₂ repulsion and coupling data with the F₃ progeny test, use Fisher's scoring method to estimate p.
   
   
9. The maximum likelihood method uses calculus to find the value of p that maximizes the probability that the observed data fits the model.