Karl Broman
Statistical Methods for the Analysis of Recombinational Processes
Recombination plays a central role in shaping the organization of variation along chromosomes on both generational and evolutionary time scales. We are studying the locations of crossovers on a single chromosome in reciprocal backcrosses, with each of female and male F1. There is a clear difference in the recombination pattern in males and females, and we further find a difference in the locations of recombination events in the reciprocal crosses in the male (but not in the female), indicating a potential imprinting effect on recombination hotspots. Particularly exciting is our work on crossover interference, and its relation to the sex difference in recombination rate. Crossover interference appears to act on a physical level; when the difference in the compaction of chromosomes is taken into account, the distances between crossovers in males and females are nearly identical. We hypothesize that crossover interference, and the difference in the compaction of chromosomes in males and females, is the primary cause of the sex difference in recombination rate. (Females have longer chromosomes at meiosis, and so may exhibit more crossovers.)
This work indicates the need for new probability models for the recombination process. In the past, crossover interference has been modeled with respect to genetic distance. That crossover interference is acting on the physical level implies that it should be modeled with respect to physical distance, but this requires a probability model that accounts for both interference and the variation in recombination rate along the chromosome. (The latter effect could be ignored when genetic distance was considered.) Pair-wise interaction models appear to be most appropriate for our purposes, but there are a number of technical difficulties in fitting such models to data, which will need to be overcome.
