The mode of within-locus gene action in most genomic regions is termed
as the major genomic mode, i.e., it is the within-locus allelic effec
ts in most regions of the genome. Determining whether dominance or ove
rdominance is the major genomic mode is important for two long-standin
g evolutionary genetics issues: 1. How is the genetic variation in mos
t genomic regions maintained? 2. What is the major mechanism for heter
osis? Many efforts have been made, but almost all of them suffer some
explanational difficulties. Here we propose an alternative inference a
pproach. It is based on the existent theoretical results on the correl
ation of the recombination rate and the level of neutral variation in
different genomic regions. Positive and negative correlations suggest
dominance and overdominance, respectively, as the major genomic mode.
Zero correlations imply either few selected sites or about equal compo
sition and distribution of dominant and overdominant regions in the ge
nome, depending on the data distribution. This approach not only avoid
s all the problems associated with earlier approaches, but it is also
particularly useful in organisms where controlled breeding is difficul
t. Well-corroborated data in Drosophila and recently emerging data in
mice and humans all suggest dominance as the major genomic mode.