haplotype inference

Harrigan RJ, Mazza ME & Sorenson MD 2008 Computation vs. cloning: evaluation of two methods for haplotype determination. Mol Ecol Res 8:1239-1248.

• the computational method may actually outperform cloning, at least when the latter is based on a limited number of clones derived from PCR products generated with standard Taq DNA polymerase

• increasing sample size is a double-edged sword
• additional unique polymorphisms (n = 6) were encountered in the larger data set

• PHASE performed well despite some level of natural recombination and homoplasy
• extra steps in the haplotype network (Fig. 1) are concentrated near the ends of our sequences
• suggesting the occurrence of natural recombination events

• most of the problems we encountered with cloning might have been avoided through the use of a high fidelity polymerase such as Pfu