A simple method for the estimation of recombination frequencies and genetic distances

We have developed an alternative approach to the maximum likelihood method for calculating recombination values and linkage intensities. This new method is simpler than those in current use in that it obviates the need for formulae and tables. Maximum likelihood methods imply the use of iterative procedures over highly complicated estimation equations (at least second degree polynomials), whereas simplified methods use single-step procedures involving simple linear equations. The proposed method uses the frequencies of the distinguishable phenotypes that came from the union of at least one recombinant gamete with another gamete. We performed Monte-Carlo simulations for various combinations of genetic distance and offspring size. The recombination values obtained using the new method were compared with those derived by the maximum likelihood method on both simulated and experimental data. They were found to be nearly identical. The observed distribution of the recombination frequency values does not differ significantly from the Normal distribution, except for extreme values of the mean, as the Skewedness and kurtosis coefficients do not differ from zero. Our method has a similar accuracy to the maximum likelihood methods for recombination frequencies smaller than 25 cM and the difference does not increase greatly for greater frequencies.