Biologists currently estimate that the smallest life form as we know it would have needed about 256 genes. (See Proceedings of the National Academy of Sciences Volume 93, Number 19, pp. 10268-10273 at http://journals.at-home.com/get_doc/1854083/8551). A gene is typically 1000 or more base pairs long, and there is some space in between, so 256 genes would amount to about 300,000 bases of DNA. The deoxyribose in the DNA ``backbone'' determines the direction in which it will spiral. Since organic molecules can be generated in both forms, the chance of obtaining all one form or another in 300,000 bases is one in two to the 300,000 power. This is about one in 10 to the 90,000 power. It seems to be necessary for life that all of these bases spiral in the same direction. Now, if we imagine many, many DNA molecules being formed in the early history of the earth, we might have say 10 100 molecules altogether (which is really much too high). But even this would make the probability of getting one DNA molecule right about one in 10 to the 89,900 power, still essentially zero. And we are not even considering what proteins the DNA generates, or how the rest of the cell structure would get put together! So the real probability would be fantastically small.
Biologists are hypothesizing some RNA-based life form that might have had a smaller genome and might have given rise to a cell with about 256 genes. Until this is demonstrated, one would have to say that the problem of abiogenesis is very severe indeed for the theory of evolution.