Using data from the HapMap project Chen and Rajewsky compared the density of single nucleotide polymorphisms (SNPs) in computationally predicted microRNA binding sites. SNPs are small changes in the nucleotide sequence that may or may not have big effects. If evolution likes a gene just the way it is then it will do its best to suppress polymorphism, so low SNP density is a hint toward negative selection. But low SNP density could also just mean a low mutation rate. Apparently, there is a technique referred to as derived allele frequency (DAF) analysis that gets around this by comparing the human polymorphisms with the ancestral sequence (chimp or macaque). I don't understand DAF analysis. Maybe someone else can explain it. Point is, this is evidence for negative selection on microRNA binding sites in humans.

They also found evidence for a SNP that was recently (within the last 50,000-70,000) positively selected in Africans. The SNP they identified is present in 87% of Africans and almost completely absent in Europeans and Asians. It is in the miRNA binding site for Map1lc3b, more popularly known as MAP1B (microtubule-associated protein 1B). MicroRNAs generally repress synthesis of whatever protein they bind. The authors don't mention whether this SNP is likely to increase or decrease the stability of binding between the miRNA and the MAP1B-RNA. They do mention that misregulation of this gene has been linked to disorders such as giant axonal neuropathy and fragile X syndrome. The protein disrupted in fragile X is itself a translational suppressor, so there could be some synergy here in the fragile X phenotype.

Microtubule associated proteins often stabilize the microtubules. These are long filaments made up of polymerized tubulin subunits. Microtubules are the cell's molecular highway. Little semis made of kinesin and dynein proteins traverse up and down the microtubule highway and carry cargo such as dendritically targeted RNAs to their destination. So one function of this SNP might be to subtly alter how many lanes the highway has at some overpasses. On the other hand, according to the abstract below, MAP1B affects neural development as well. Guess it is still way too early to say.

Microtubule-associated protein 1B function during normal development, regeneration, and pathological conditions in the nervous system. J Neurobiol. 2004 Jan;58(1):48-59.
Gonzalez-Billault C, Jimenez-Mateos EM, Caceres A, Diaz-Nido J, Wandosell F, Avila J.

Microtubule-associated protein 1B is the first MAP to be expressed during the development of the nervous system. Several different approaches have revealed that MAP1B function is associated with microtubule and actin microfilament polymerization and dynamics. In recent years, the generation of molecular models to inactivate MAP1B function in invertebrates and mammals has sparked some controversy about the real role of MAP1B. Despite discrepancies between some studies, it is clear that MAP1B plays a principal role in the development of the nervous system. In this article, we summarize the evidence for MAP1B function in a wide variety of cellular processes implicated in the proper construction of the nervous system. We also discuss the role of MAP1B in pathological processes.