You have it bass ackwards as usual Bill. If you don’t know how many possible working variations there are your method will consistently overestimate your calculated “FI”. One of the many reasons it’s such a meaningless and useless metric.
2 Likes
If I go outside the e coli strain the alignment drops radically. 18% identity for 3 different bacteria types.
Can you support this claim with real math?
In gpuccio’s FI calculations as the number of possible functioning sequences goes up the calculated FI goes down. Are you now agreeing gpuccio’s calculations are bogus?
1 Like
Where does the length of time a protein has been in existence factor into your / gpuccio’s FI calculations Bill? Where does that variable get plugged into the FI equation?
1 Like
And relatedly, how long must we wait, and why that long?
1 Like
Do you understand your statement is false? Gpuccio’s calculation is an estimate. It does not consider other possible solutions in sequence space that could solve the problem but is there any reason to think this the number is significant? In certain cases it might be but in testing this so far in most cases we see a low FI number.
On the other hand it reduces the number to a more conservative estimate by not considering all substitutability.
I understand all you can do is make excuses while not showing it is false.
Where does the length of time a protein has been in existence factor into your / gpuccio’s FI calculations Bill? Where does that variable get plugged into the FI equation?
1 Like
Rupe and Sanford show that as far as the LTEE is concerned, evolution goes the wrong way, ie., down, not up. This is pretty important.
Now, I will sincerely be happy to know what in your view are the more important findings.
Evolution doesn’t have a direction. Anyone who thinks it does doesn’t know what he/she is talking about.
1 Like
It’s a simple logic error. You are saying an unknown variable is greater or less than a known one. How do you know that the other sequences that we cannot measure are greater or less then the reduction in FI we are making by assigning 0 substitutability of unaligned sequences.
Evolution doesn’t have a way it’s going. It’s not going up, and it’s not going down. What is “up”? Up towards what? What is “down”? Down towards what? What are these directional measures supposed to be? What the heck are you talking about?
On your part. You don’t know the possible number of functioning sequences so you assume the ones we see are the only ones possible. It’s called the Sharpshooter fallacy and it’s the same blunder you’ve been making for years.
1 Like
By request, changing back to the default 3-days from last post timer.
BUT let’s try to clean this up a bit. MORE substantive comments, LESS one-liners. If you AGREE on a point, SAY SO, so there can be a little less re-hashing. STATE your own position, and your understanding of the other guy’s position. THEN work to find more points of agreement.
/fnord
How long is long enough for a real test? Why?
I would just take the preserved sequences and calculate their length.
What does that mean specifically? “Take the preserved sequences and calculate their length”? Give an example.
If it is 90% that would be about 10^150 or 500 bits close enough to your calculation.
If, what, length(?) is “90%” that would be 10^150 or 500 bits? Bill can you please explain what the fork you’re doing?
On the contrary, this has been a very substantive and revealing discussion. It has been clearly and extensively demonstrated that @gpuccio’s claims are completely without merit, and he is not able to defend the model he is suggesting. Moreover, it is also clear that those who believe his argument had merit only do so because they fail to grasp very basic and fundamental aspects of the evolutionary theory.
This is important information to have, to the extent that anything dealing with the issue of whether ID arguments hold up can be deemed “important.” That some members are unable or unwilling to accept these conclusions does not diminish their importance.
1 Like
You need time for mutations to accumulate and fix in the population. The longer the better. Saturation happens in about 400 million years according to gpuccio.
I simply would shorten the sequence by eliminating non preserved AA sequences. So if the number of AA’s is 100 and the conservation is 90% then I would take 20^90 as the functional space. This does not consider the same function in a very different sequence so the number is an estimate and assumes that the other functional sequence space is not significant. I believe a reasonable error bound should cover the additional functionality possibility.
90% or 130 is 120 AAs preserved 20^120 is about 10^150. I have not bothered to do precise calculation here.
Wrong. You don’t. This is one of the basic things you do not understand.
Do you care to explain?
The same goes for ubiquitin. If you move outside of eukaryotes the identity drops radically.
Do you really think that brand new protein with brand new function that evolves in the lab is going to suddenly show up in multiple bacterial species? If not, why are you using the genomes of other species?
You have asked for examples of high FI proteins evolving in the lab in real time, but then you turn around and require it to be a widely shared protein with 600 million years of evolutionary history. You need to make up your mind.
2 Likes