Joining genetic linkage maps using a joint likelihood function

• DOI: 10.1007/s00122-004-1705-x
• Publication Date: 2004-06-24T09:45:30Z
• Authors: Xin-Sheng Hu; Carol Goodwillie; Kermit M. Ritland
• Keywords: 0301 basic medicine; Likelihood Functions; Magnoliopsida; 0303 health sciences; 03 medical and health sciences; Phenotype; Chromosome Mapping; Computer Simulation; Nucleic Acid Amplification Techniques; Crosses, Genetic; Polymorphism, Restriction Fragment Length; Software

We present an efficient method to join genetic maps derived from different crosses, which is especially appropriate for dominant markers. In contrast to the "JoinMap" algorithm, which estimates information about recombination in a given cross from the LOD values and then combines estimates among crosses assuming a binomial sampling distribution, we construct a joint likelihood function that combines information across all crosses, to obtain a joint estimate of recombination. Simulations indicated that the difference between these two approaches is small when codominant markers are used, but that the joint likelihood approach shows substantially improved estimates when dominant or a mixture of dominant and codominant markers are used. This is because the joint likelihood implicitly finds the optimal weights among different classes of data, while the former method does not accurately predict the information from crosses involving dominant markers. Application of our method is illustrated by the construction of a linkage map for Linanthus using both backcrosses and the F2 of a cross between Linanthus jepsonii and L. bicolor, assayed with amplified fragment length polymorphisms (AFLP).