One of the common claims heard in these debates is that there is a low probability of random sequence having function. Catalytic antibodies may offer a good way of testing this claim.
During embryonic development, B-cells go through a process where segments of DNA are randomly shuffled and stitched together (VDJ recombination) to make the variable regions of antibodies, the sections of the antibody that are responsible for sticking to bacteria, viruses, and other antigens. Each B-cell lineage has just one combination of this shuffled DNA, and there are about 100 million B-cell lineages in each human being and in other mammalian species [NOTE: I’m not sure how this may or may not work in more distantly related vertebrate species, so I am sticking with mammals].
So what are the chances that these randomly shuffled bits of DNA will have enzyme activity? As it turns out, antibodies with enzyme function are relatively common. The following review article discusses several different catalytic antibodies, one of which is pretty crazy:
The existence of catalytic antibodies tells me that evolving function is relatively easy. Shuffling a few bits of DNA around produces many different and specific catalytic enzymes, some of which include beta-lactamase activity which some ID supporters have argued is almost impossible to evolve.