The BSC4 gene in Saccharomyces cerevisiae, or baker’s yeast, was interesting to researchers because of its leaky stop codon but then it became more interesting because it only showed up in that one particular organism. The BSC4 gene is yet another example of a species-specific, or de novo, protein-coding gene. In this case, the protein appears to be involved in DNA repair and helping the organisms cope with nutrient-poor environments. And if this protein is anything like a typical protein, then evolution, even by the evolutionist’s own reckoning, would not be able to construct such low-probability designs. The BSC4 gene must have been constructed in only the past 10 million years or so. And evolutionists cannot appeal to speculative mechanisms. No exon shuffling, duplication, retroposition, fusion, fission or whatever for this gene. It must have arisen the old fashion way, by an evolutionary search through sequence space via random mutations. Is this feasible? The important numbers here are the number of attempts that are possible, and the number of attempts that are required. The two might not add up.
Evolutionists have estimated the number of attempts that evolution could possibly have to construct a new protein. Their upper limit is 10^43 (a one followed by 43 zeros) obtained by multiplying 10^30 (cells in the world) by 10^4 (new genes generated per cell per year) by 10^9 (years). The lower limit is 10^21 obtained by multiplying 10^9 (bacteria species in the world) by 10^3 (unique sequences per species) by 10^9 (years).
While these estimates are incredibly optimistic for several reasons (even the evolutionists admit the upper limit is an “extreme upper limit”), we’re going by the evolutionist’s numbers for the moment. They need to be adjusted, however, because those estimates are for the entire history of the world and for all the species. For de novo genes such as BSC4, we’re dealing with a single species (with an effective population size of about ten million), and only about ten million years. In this case the upper and lower limits become 10^18 and 10^10, respectively.
And the numbers are even smaller for de novo genes found in humans. The time allowed goes down to about 5 million years and the effective population size goes down by at least two orders of magnitude, to about 10^5. So in this case the upper and lower limits become 10^14 and 10^10, respectively.
Studies have shown that even this many searches does not produce much. One study found that with 10^12 attempts all that was produced were a few proteins with weak ATP binding.
And for typical proteins, even these optimistic estimates of the number of attempts fall short by more than 50 orders of magnitude. And these deficits are according to the evolutionist’s own estimates of how many attempts would be required to find a typical protein. One study concluded that 10^63 attempts would be required for a relatively short protein. And a similar result (10^65 attempts required) was obtained by comparing protein sequences.
Another study found that 10^64 to 10^77 attempts are required, and another study concluded that 10^70 attempts would be required. In that case the protein was about the same length as that encoded by BSC4, but it was only a part of a larger protein which otherwise was intact, thus making the search easier.
The numbers don’t add up. The evolution of de novo genes can only count on from 10^10 to 10^18 attempts (and that’s optimistic). If the new proteins are anything like typical proteins, then these numbers show astronomical problems.