[stitchy]

"Our model makes two very strong predictions. The first is that the rate of de novo mutations in affected children from multiplex families will be barely above that of an unaffected control population. On this point, the literature presently has conflicting reports [3,6,22]. The samples needed to resolve this question properly, namely blood derived DNAs from multiplex families, has either not yet been collected or sequenced. The second prediction made by our model is that there will be a high rate of transmission of strong alleles in multiplex families, typically from the mother, and in the same genes targeted by de novo mutation in simplex autism. While there is some indirect support for a female carrier effect based on half-sibs [23], a genetic study of transmission based on the targets of de novo mutation is only now possible [21]."

This seems like a succinct summary for the paper. Or at least, it's the part that stuck out to me.