Via Pharyngula, some non-scientist MD thinks that ~21,000 protein-coding genes aren't nearly enough to make a human (which of course then means that evolution is wrong):

4) The Human Genome Project showed that only 1-2% of Human DNA codes for proteins, or about 25,000 genes. These are not enough to account for the complexity of the organism. What is the other 98% of the genome's function? We don't know.

(BTW, the count of human genes has gone down since the genome sequence was first released; the latest number I hear from my gene-finding colleagues is about 21,000.)

PZ points out the absurdity of this claim that we're short on genes:

This is fast becoming one of the most popular assertions by creationists. Exactly how many genes would be sufficient to account for the complexity of a human being? Show your work. How many genes do we need to have to make you happy, and why should your sense of self-worth be a reason for us to have more?

It's true that before the human genome was sequenced people estimated that humans had about 100,000 genes, but the truth is, scientists had no clue, and 100,000 was just a guess, not based on some theory of how many genes are required for a given amount of complexity. It was a guess that turned out to be spectacularly wrong. But the non-scientist MD seems to think that we don't really understand human developmental biology, and that there must be some mysterious stuff in the bulk of the non-protein-coding part of the genome to account for our supposed lack of sufficient numbers of protein-coding genes.

Let's ignore for a minute the absurdity of making vague claims about how many genes would be enough to make a human being, and ask a slightly different question which is maybe less vague: given that it takes about 16,000 genes to make a fruit fly, how many more genes would we expect humans to have?

The answer is not much, which is counter intuitive if you are used to thinking of humans as the pinnacle of evolutionary complexity (or the pinnacle of creation). But on a cellular or molecular level, fruit flies are not that different from (or much less complex than) humans, which is why fruit flies make such great model research organisms. Fruit flies have many of the same basic tissue types that we have, such as nerve, muscle, fat, gut, and gonadal cells. The largest source of differences between humans and fruit flies is not our molecular components; it's how those components are used. In fact, I'm not sure that it would be absolutely necessary to have the extra 5,000-6,000 genes to make a more 'complex' organism like us. It is probably possible engineer or evolve a human-like organism with about the same number of genes found in a fruit fly.

The bottom line is that, if we take it as given that you can make a fruit fly with 16,000 genes, there is no reason to think that humans don't have enough protein-coding genes. There is no need to speculate about a mysterious source of information in the rest of the genome. Some of that DNA is regulatory DNA, some codes for RNA genes, and most of it is probably functionally neutral. Even if there is some new mechanism of regulation hidden in the genome, it's relative significance will be small: the bulk of the developmental heavy lifting is done by protein-coding genes and their associated regulatory sequences, with a smaller role played by non-coding RNAs.