The odds of a random mutation causing amino acids to line up in a chain approximately 140 long to make one of these functional proteins is 1 out of 1 with 40 zeros after it. By comparison the odds of winning the lottery is only 1 out of 1 with 8 zeros after it. It is ridiculous to think this is the best explanation for how this happened.
Well, I have some time so I'll go over some reasons creationist probabilities are wrong. And this is creationist math - I've seen dozens, if not hundreds, of creationists making these same silly calculations in the course of around thirty years lurking on various debate forums. ZNP didn't show his work, but here are some common implicit assumptions and omissions.
1. The protein has a fixed length
2. There is only one possible sequence that works
3. there are 20 possible amino acids
4. All other types of mutations are ignored.
5. Post hoc calculations are meaningless
1. Yes, you have to start somewhere - but the probability of getting a functional protein might be quite dependent on the length of the protein.
2.This is a big one: no one who has the least familiarity with biochemistry would think this is true. Looking at the differing sequences of proteins from various species - all with identical functions - makes one realize that there are a huge number of different sequences that can perform any particular function. So you need to calculate the percentage of possible sequences that work. You might be able to find various estimates of that percentage, but you can't simply calculate it - it needs to be measured. That is why I tried (somewhat unsuccessfully, I'll admit) to provide ZNP with empirically measured percentages. Of course, there is no reason to think that there is a single percentage - some functions might be much harder to achieve than others.
3. Not necessarily relevant, but there were quite possibly fewer amino acids utilized early on in the history of life on earth
4. This is also a very big one: no one who has a good familiarity with evolutionary science thinks that forming proteins from random sequences is the most common path to novel sequences. Duplications of existing proteins means that you start with a protein that folds correctly and already has a function. Further mutations can make modifications to that function. This is amply supported by evidence that duplicated genes make up large swaths of our functional genome.
5. I'll let other posters have fun with this. If you make post hoc calculations on any set of mostly random events, you might conclude they are impossible.