# Comments on Gpuccio: Functional Information Methodology

**URL:** <https://discourse.peacefulscience.org/t/comments-on-gpuccio-functional-information-methodology/7560>\
**Category:** Side Conversation\
**Tags:** Comments\
**Created:** [August 26, 2019, 3:59pm UTC](https://discourse.peacefulscience.org/t/comments-on-gpuccio-functional-information-methodology/7560 "2019-08-26T15:59:38Z")\
**Posts on this page:** 20\
**Page:** 16

<div class="post-metadata">

**Author:** ![colewd](https://avatars.discourse-cdn.com/v4/letter/c/ecd19e/32.png) [@colewd](https://discourse.peacefulscience.org/u/colewd)\
**Post date:** [August 31, 2019, 8:16pm UTC](https://discourse.peacefulscience.org/t/comments-on-gpuccio-functional-information-methodology/7560/364 "2019-08-31T20:16:41Z")

</div>

> [@swamidass](#):
>
> A is functional  
> B is a different function  
> A + B is a different function still.
> 
> That is pretty clear. that is the definition of decomposability.

Are you defining the function of A + B here? Can we call this C that meets HS 🙂

---

<div class="post-metadata">

**Author:** ![swamidass](https://sea2.discourse-cdn.com/flex016/user_avatar/discourse.peacefulscience.org/swamidass/32/3_2.png) [@swamidass](https://discourse.peacefulscience.org/u/swamidass)\
**Post date:** [August 31, 2019, 8:18pm UTC](https://discourse.peacefulscience.org/t/comments-on-gpuccio-functional-information-methodology/7560/365 "2019-08-31T20:18:04Z")

</div>

The function of A+B is:

Either, merely function A and B together.

Or, function A and B together, with some new function C that arises from those two.

It does not really matter though. If you have A and you have B, **by definition** of function C, you have C.

---

<div class="post-metadata">

**Author:** ![colewd](https://avatars.discourse-cdn.com/v4/letter/c/ecd19e/32.png) [@colewd](https://discourse.peacefulscience.org/u/colewd)\
**Post date:** [August 31, 2019, 8:22pm UTC](https://discourse.peacefulscience.org/t/comments-on-gpuccio-functional-information-methodology/7560/366 "2019-08-31T20:22:52Z")

</div>

I think that works The function A+B=C. So for HS you are defining function C and the bits that make up A and B. How those bits came about is a separate issue.

Waiting time being improved by adding resources is certainly true. I think @gpuccio would agree with you and Neil.

---

<div class="post-metadata">

**Author:** ![Dan\_Eastwood](https://sea2.discourse-cdn.com/flex016/user_avatar/discourse.peacefulscience.org/dan_eastwood/32/2368_2.png) [@Dan\_Eastwood](https://discourse.peacefulscience.org/u/Dan_Eastwood)\
**Post date:** [August 31, 2019, 9:10pm UTC](https://discourse.peacefulscience.org/t/comments-on-gpuccio-functional-information-methodology/7560/367 "2019-08-31T21:10:09Z")

</div>

> [@Ashwin\_s](#):
>
> The situation is a little more dynamic that this. The environment/ecosytem changes (change in weather, animal migration patterns, pathogens etc).
> 
> So it more like the safes code keeps changing every once in a while.  
> This is why evolution is a process that depended so much on “contingencies”.  
> In short biodiversity is a miracle that normally shouldn’t have happened.

The safes example avoids those contingencies, and yes there are contingencies, but contingency is also opportunity for change. It the thief gets stuck on one lock, perhaps another will open.

It’s not a bit like changing combinations. **The combinations never change.** The amount of money in each safe can change (biological reward), and new sets of safes may appear (or be removed) by contingency, but the sequence to unlock each safe should never change. Changing the sequence would be equivalent to changing the physical laws that govern chemistry - those do not change.  
The thief might forget a combination, or forget to pass the secret on to his grandson (loss of function), but the biological function the combination represents in this example cannot change.

---

<div class="post-metadata">

**Author:** ![Mercer](https://avatars.discourse-cdn.com/v4/letter/m/e274bd/32.png) [@Mercer](https://discourse.peacefulscience.org/u/Mercer)\
**Post date:** [August 31, 2019, 9:13pm UTC](https://discourse.peacefulscience.org/t/comments-on-gpuccio-functional-information-methodology/7560/368 "2019-08-31T21:13:35Z")

</div>

> [@Giltil](#):
>
> I will first take a conservative approach and assume that in order to produce the function, a specific mutation (a specific aa change) must occur at a specific position in the V region of a B cell receptor within a single activated B cell.

You are taking a polemic approach, not a conservative one.

You have no basis for claiming that this is about a single mutation. You’re ignoring the number of different clones (and the number of different branches of the same clone) in which this is evolution is happening simultaneously in a typical vertebrate.

If you were being conservative and not polemic, you would have included factors for both of these well-documented realities. We have sequencing data. Why don’t you look at some before taking a position?

> [@Giltil](#):
>
> Let’s call p this specific location and r this specific mutation. Now let’s define the following 3 probabilities P1, P2 and P3, with:
> 
> P1: the probability that a mutation occurs in the V region of a B cell receptor during a cell division. (I’ve found in my immunology textbook that P1 is close to ½)

You are making the false assumption that there is only a single cell division. We know that the average clone (germinal center) is about 10000 cells and the doubling time is about 6 hours.

If you were looking for the reality, you would consider the number of cell divisions, but that would give an FIb far too high for your desired conclusion.

> [@Giltil](#):
>
> P2: the probability that the above mutation occurs at position p in the V region (given that the V region of immunoglobulins is 120aa long, we have P2=1/120)

There is no reason to pretend that only one position is involved, as mutations in many different positions have been shown to increase affinity. People have done microsequencing of immunoglobulin genes from microsamples taken from different parts of single germinal centers. Tens of thousands of monoclonal antibodies have been made. The data from those studies show your assumption to be false.

> [@Giltil](#):
>
> P3: the probability that the above mutation that occurred at position p corresponds to r (given that there are 20 different aa, and assuming for simplicity that each aa change is equiprobable, we have P3=1/20)

Since you’ve ignored many factors that significantly increase the probability, your estimated probability is wrong.

> [@Giltil](#):
>
> The bottom line is that the amount of new FI produced by the immune system is, to say the least, ridiculously low compared to the 500 bits threshold required to draw a design inference.

The bottom line is that your estimate of the amount of new FIb produced (you still haven’t estimated FIa) should be far higher than your ridiculously low polemic estimate.

> [@Giltil](#):
>
> In fact, I have revised my estimation of FIb and now think it is lower, around 50 bits.

Since you offered no reasons for doing so, it appears all but certain that you have revised it in hindsight because you have to lower it to preserve @gpuccio’s claim. As shown above, your estimate has little basis in reality. You’ve ignored multiple factors that increase both the information content and the probability of mutations that increase it.

I also haven’t seen an estimate of FIa. How come?

> [@Giltil](#):
>
> The question now is why the RV + NS is able to produce high FI in a few weeks during the process of somatic hypermutation (SHM).

We know a lot about why that is and you are ignoring the majority of what we know. We know enough to know that your estimate is wishful thinking and that @gpuccio’s claim is false.

> [@Giltil](#):
>
> So it is true that RV + NS can produce high FI but only in one particular and very special situation, ie., when the final target exhibiting high FI can be reach incrementally through a serie of small selective steps.

I don’t see how a system shared by virtually all vertebrates that functions constantly throughout life can be called “special.” @gpuccio was the one who specified vertebrates, wasn’t he?

> [@Giltil](#):
>
> Moreover, the FI associated with each of these selection steps is very low, around 10 bits.

Your estimate does not address what we know about the process. You just made it up.

> [@Giltil](#):
>
> Such a very special situation is quite rare in biology and doesn’t apply to complex proteins.

Since @gpuccio specified vertebrates and this is a universal vertebrate system that functions constantly, there’s no reason, other than a need to avoid science and engage in polemics, to call something like that “special” or “quite rare.”

In fact, it should be one of the most obvious and first systems one should study if looking at the biological creation of functional information, quadruply obvious for one trained as a physician.

---

<div class="post-metadata">

**Author:** ![Rumraket](https://sea2.discourse-cdn.com/flex016/user_avatar/discourse.peacefulscience.org/rumraket/32/9328_2.png) [@Rumraket](https://discourse.peacefulscience.org/u/Rumraket)\
**Post date:** [August 31, 2019, 10:46pm UTC](https://discourse.peacefulscience.org/t/comments-on-gpuccio-functional-information-methodology/7560/369 "2019-08-31T22:46:57Z")

</div>

I’m still missing an explanation for why, if a handful of mutations can fix incrementally when antibodies are evolved to target a particular antigen, why could more mutations not fix over even longer timescales and the FI increase further then? I don’t understand what this mysterious barrier to reach 500 bits of FI is supposed to be.

---

<div class="post-metadata">

**Author:** ![Giltil](https://avatars.discourse-cdn.com/v4/letter/g/5f8ce5/32.png) [@Giltil](https://discourse.peacefulscience.org/u/Giltil)\
**Post date:** [August 31, 2019, 10:57pm UTC](https://discourse.peacefulscience.org/t/comments-on-gpuccio-functional-information-methodology/7560/370 "2019-08-31T22:57:16Z")

</div>

> [@Mercer](#):
>
> You are making the false assumption that there is only a single cell division. We know that the average clone (germinal center) is about 10000 cells and the doubling time is about 6 hours.
> 
> If you were looking for the reality, you would consider the number of cell divisions, but that would give an FIb far too high for your desired conclusion.

You are confused here. I am not at all making the false assumption that there is only a single cell division. I know perfectly well that there are many activated B cells that undergo many cell divisions. But these factors have nothing to do with FIb; rather, they pertain to the probabilistic resources of the system. And it is these probabilistic resources that explain why the system is able to find the target (an antibody with higher affinity) so easily.  
Now, if you are interested, I can give you a more thorough account of how I arrived at this value of 50 bits for FIb.

---

<div class="post-metadata">

**Author:** ![Timothy\_Horton](https://sea2.discourse-cdn.com/flex016/user_avatar/discourse.peacefulscience.org/timothy_horton/32/9410_2.png) [@Timothy\_Horton](https://discourse.peacefulscience.org/u/Timothy_Horton)\
**Post date:** [August 31, 2019, 11:11pm UTC](https://discourse.peacefulscience.org/t/comments-on-gpuccio-functional-information-methodology/7560/371 "2019-08-31T23:11:40Z")

</div>

> [@Rumraket](#):
>
> I’m still missing an explanation for why, if a handful of mutations can fix incrementally when antibodies are evolved to target a particular antigen, why could more mutations not fix over even longer timescales and the FI increase further then? I don’t understand what this mysterious barrier to reach 500 bits of FI is supposed to be.

I suspect it’s the same as the mysterious barrier Creationists claim makes it impossible for micro-evolution changes to accumulate over time into macro-evolution. 🙂

---

<div class="post-metadata">

**Author:** ![Mercer](https://avatars.discourse-cdn.com/v4/letter/m/e274bd/32.png) [@Mercer](https://discourse.peacefulscience.org/u/Mercer)\
**Post date:** [August 31, 2019, 11:20pm UTC](https://discourse.peacefulscience.org/t/comments-on-gpuccio-functional-information-methodology/7560/372 "2019-08-31T23:20:34Z")

</div>

> [@Giltil](#):
>
> You are confused here. I am not at all making the false assumption that there is only a single cell division.

I don’t think so. Your probability calculation assumed a single division and included no factors representing the numbers of cell divisions. That means that the number you calculated, if everything else in it was right (it wasn’t), would only represent the probability for a single cell division and therefore useless.

> [@Giltil](#):
>
> I know perfectly well that there are many activated B cells that undergo many cell divisions. But these factors have nothing to do with FIb; rather, they pertain to the probabilistic resources of the system. And it is these probabilistic resources that explain why the system is able to find the target (an antibody with higher affinity) so easily.

That makes no sense.

> [@Giltil](#):
>
> Now, if you are interested, I can give you a more thorough account of how I arrived at this value of 50 bits for FIb.

That would be fun. Where’s your estimate for FIa?

---

<div class="post-metadata">

**Author:** ![Giltil](https://avatars.discourse-cdn.com/v4/letter/g/5f8ce5/32.png) [@Giltil](https://discourse.peacefulscience.org/u/Giltil)\
**Post date:** [September 1, 2019, 8:48am UTC](https://discourse.peacefulscience.org/t/comments-on-gpuccio-functional-information-methodology/7560/374 "2019-09-01T08:48:13Z")

</div>

> [@Mercer](#):
>
> > P2: the probability that the above mutation occurs at position p in the V region (given that the V region of immunoglobulins is 120aa long, we have P2=1/120)
> 
> There is no reason to pretend that only one position is involved, as mutations in many different positions have been shown to increase affinity

You are perfectly right here, and this is precisely why I said I have used a conservative approach in my first attempt at computing FIb. Because if, as it is indeed the case, different mutations at different positions can increase affinity, or, put differently, if the target space increases, then FIb would decrease. Hence my use of the word « conservative ». There is nothing polemical here!

> [@Mercer](#):
>
> > In fact, I have revised my estimation of FIb and now think it is lower, around 50 bits.
> 
> Since you offered no reasons for doing so, it appears all but certain that you have revised it in hindsight because you have to lower it to preserve @gpuccio’s claim. As shown above, your estimate has little basis in reality. You’ve ignored multiple factors that increase both the information content and the probability of mutations that increase it.

The reason I have revised my estimation of FIb is precisely to take into account that more than one mutation can increase affinity. Are you going to blame me for trying to be more accurate in my estimation of FIb?

---

<div class="post-metadata">

**Author:** ![Giltil](https://avatars.discourse-cdn.com/v4/letter/g/5f8ce5/32.png) [@Giltil](https://discourse.peacefulscience.org/u/Giltil)\
**Post date:** [September 1, 2019, 10:18am UTC](https://discourse.peacefulscience.org/t/comments-on-gpuccio-functional-information-methodology/7560/375 "2019-09-01T10:18:35Z")

</div>

> [@Mercer](#):
>
> > I know perfectly well that there are many activated B cells that undergo many cell divisions. But these factors have nothing to do with FIb; rather, they pertain to the probabilistic resources of the system. And it is these probabilistic resources that explain why the system is able to find the target (an antibody with higher affinity) so easily.
> 
> That makes no sense.

It makes no sense for you because you don’t understand the basic concepts used by IDers. Let me try to explain. As @gpuccio defined it, FI is a measure of the improbability of finding the target space in one random event. The important point here is the expression « in one random event ». This is why when computing FIb, you have first to compute the probability that a single B cell will produce a receptor with a higher affinity through a single cell division. I have estimated this probability to be 1/1000, which corresponds to a FI of 10 bits. If, in order to produce this amount of FI, the immune system had to rely on only one B cell division, ie., was only allowed to perform a single try, the somatic hypermutation (SHM) process would not work. But why is it that it works? The answer is that during the SHM process, many different B cells are allowed to undergo many cell divisions, which considerably increase the probability of the system to find the target. IOW, the system works because it allows a large number of try.  
Or, put differently, the system works because its probabilistic resources are important.

Let me now give you a non biological example of these concepts.  
Imagine you have a random letter generator set to output a sequence of 3 letters.  
And imagine you want it to produce the word « cat ».  
In order to compute the FI associated to the word « cat » in this situation, you will have to calculate the probability to produce the word « cat » through a SINGLE try. This probability is equal to 1/26^3, that is 1/17576, which corresponds to an FI of 14 bits.  
Now, if you are allowed to play that game not once but many times (if the probabilistic resources at hand increase) you will obviously enhance your chance to find the target. IOW, the more the probabilistic resources, the more the chance of finding the target. But the important point to see here is that the probabilistic resources have nothing to do whatsoever with the FI. The FI associated to the word « cat » in this example remains the same (14 bits) whatever the number of try. I hope things are clearer for you now.

---

<div class="post-metadata">

**Author:** ![Rumraket](https://sea2.discourse-cdn.com/flex016/user_avatar/discourse.peacefulscience.org/rumraket/32/9328_2.png) [@Rumraket](https://discourse.peacefulscience.org/u/Rumraket)\
**Post date:** [September 1, 2019, 10:47am UTC](https://discourse.peacefulscience.org/t/comments-on-gpuccio-functional-information-methodology/7560/376 "2019-09-01T10:47:12Z")

</div>

> [@Giltil](#):
>
> And imagine you want it to produce the word « cat ».  
> In order to compute the FI associated to the word « cat » in this situation, you will have to calculate the probability to produce the word « cat » through a SINGLE try. This probability is equal to 1/26^3, that is 1/17576, which corresponds to an FI of 14 bits.  
> Now, if you are allowed to play that game not once but many times (if the probabilistic resources at hand increase) you will obviously enhance your chance to find the target. IOW, the more the probabilistic resources increase, the more the chance of finding the target increase. But the important point to see here is that the probabilistic resources have nothing to do whatsoever with the FI. The FI associated to the word « cat » in this example remains the same (14 bits) whatever the number of try. I hope things are clearer for you now.

So why can’t a complex protein evolve?

---

<div class="post-metadata">

**Author:** ![Giltil](https://avatars.discourse-cdn.com/v4/letter/g/5f8ce5/32.png) [@Giltil](https://discourse.peacefulscience.org/u/Giltil)\
**Post date:** [September 1, 2019, 3:33pm UTC](https://discourse.peacefulscience.org/t/comments-on-gpuccio-functional-information-methodology/7560/377 "2019-09-01T15:33:24Z")

</div>

> [@Mercer](#):
>
> > Now, if you are interested, I can give you a more thorough account of how I arrived at this value of 50 bits for FIb.
> 
> That would be fun

Ok. Below is a more complete account of how I have tried to assess FIb.

When a foreign antigen enter the body, the first phase of the humoral response is the recognition of the antigen by some B cells patrolling the human body that happen to express a receptor with, most often, a weak affinity for the antigen. The second phase occurs when these activated B cells migrate into follicles and proliferate to form germinal centers where they undergo V-region somatic hypermutations (SHM) that improve affinity for the antigen. Since several rounds of selection occurs during this SHM phase, the FI produced at each round must be estimated in order to calculate the FI produced by the whole SHM process.

So let’s look first at the new FI produced during the first round of selection. Here, we first have to define a function and then calculate the probability for the organism to find this function. The function for the activated B cells can be defined as follow: « bind the antigen with higher affinity ». Let’s call P the probability to find that function.

I have first taken a conservative approach and assumed that in order to produce the function, a specific mutation (a specific aa change) must occur at a specific position in the V region of a B cell receptor within a single activated B cell. Let’s call L this specific location and R this specific mutation. Now let’s define the following 3 probabilities P1, P2 and P3, with:

P1: the probability that a mutation occurs in the V region of a B cell receptor during a cell division. (I’ve found in my immunology textbook that P1 is close to ½)

P2: the probability that the above mutation occurs at position L in the V region (given that the V region of immunoglobulins is 120aa long, we have P2=1/120)

P3: the probability that the above mutation that occurred at position L corresponds to R (given that there are 20 different aa, and assuming for simplicity that each aa change is equiprobable, we have P3=1/20)

Now given that P=P1xP2xP3, we have P=1/4800.

But this is undoubtedly an underestimation of P because in my calculation above, I have wrongly assumed that the function could only be achieved by a single specific mutation occurring at a specific position of the V region. IOW, I have assumed that the target space was equal to 1, which is obviously not the case. I don’t know the size of the target space here, but my educated guess would be around 5. In that case, we arrive at P=1/1000. Thus, the FI produced during the first round of selection of the SHM phase is equal to 10 bits.

Now, in order to calculate FIb, that is the total amount of FI produced during the SHM process, we must know what is the number of selection rounds that occurs during that process. Here again, I don’t know the answer, but a mean of 5 would be my educated guess. In that case, assuming that the same level of FI is produced at each round, we arrived at FIb=10x5, ie., 50 bits.

I am well aware of at least some of the limitations and shortcuts in my calculation of FIb but I think it is not too far from the truth.  
If someone here is willing to offer another better estimation, that would be great.

---

<div class="post-metadata">

**Author:** ![Timothy\_Horton](https://sea2.discourse-cdn.com/flex016/user_avatar/discourse.peacefulscience.org/timothy_horton/32/9410_2.png) [@Timothy\_Horton](https://discourse.peacefulscience.org/u/Timothy_Horton)\
**Post date:** [September 1, 2019, 3:36pm UTC](https://discourse.peacefulscience.org/t/comments-on-gpuccio-functional-information-methodology/7560/378 "2019-09-01T15:36:37Z")

</div>

> [@Giltil](#):
>
> It makes no sense for you because you don’t understand the basic concepts used by IDers. Let me try to explain. As @gpuccio defined it, FI is a measure of the improbability of finding the target space in one random event. The important point here is the expression « in one random event ».

Explain to us again why this is relevant since no one in science says or thinks proteins arose through one random event?

---

<div class="post-metadata">

**Author:** ![Rumraket](https://sea2.discourse-cdn.com/flex016/user_avatar/discourse.peacefulscience.org/rumraket/32/9328_2.png) [@Rumraket](https://discourse.peacefulscience.org/u/Rumraket)\
**Post date:** [September 1, 2019, 3:58pm UTC](https://discourse.peacefulscience.org/t/comments-on-gpuccio-functional-information-methodology/7560/379 "2019-09-01T15:58:00Z")

</div>

> [@Giltil](#):
>
> But this is undoubtedly an underestimation of P because in my calculation above, I have wrongly assumed that the function could only be achieved by a single specific mutation occurring at a specific position of the V region. IOW, I have assumed that the target space was equal to 1, which is obviously not the case. I don’t know the size of the target space here, but my educated guess would be around 5. In that case, we arrive at P=1/1000. Thus, the FI produced during the first round of selection of the SHM phase is equal to 10 bits.
> 
> Now, in order to calculate FIb, that is the total amount of FI produced during the SHM process, we must know what is the number of selection rounds that occurs during that process. Here again, I don’t know the answer, but a mean of 5 would be my educated guess. In that case, assuming that the same level of FI is produced at each round, we arrived at FIb=10x5, ie., 50 bits.

This all makes sense to me, the problem is I don’t see how this could in any way be taken to imply there is some sort of barrier to evolving any known protein.

What is missing from Gpuccio, or you, or anyone, is some sort of argument that connects this calculation of FI to a known protein sequence, which concludes that therefore this known protein sequence could not have evolved(or is at least extremely implausible to have evolved), so must have been designed instead.

What we need is the demonstration that there are proteins that actually have so many bits of FI that their evolution becomes implausible.

---

<div class="post-metadata">

**Author:** ![colewd](https://avatars.discourse-cdn.com/v4/letter/c/ecd19e/32.png) [@colewd](https://discourse.peacefulscience.org/u/colewd)\
**Post date:** [September 1, 2019, 4:40pm UTC](https://discourse.peacefulscience.org/t/comments-on-gpuccio-functional-information-methodology/7560/380 "2019-09-01T16:40:28Z")

</div>

> [@Rumraket](#):
>
> What we need is the demonstration that there are proteins that actually have so many bits of FI that their evolution becomes implausible.

Let’s start from the position that his calculation may have error due to not measuring all possibilities. It got stuck in some optimal position through mutation and selection or drift. The issue is still real for long proteins as even 50% substitutability for prp8 would yield 2335 bits.

This problem is so large when you take into account all the machines in the eukaryotic cell that require coordinated function the implausibility by mutation and selection is an accurate description.

You can always imagine a way it could happen but if we look at what the evidence is showing us there is almost no way to make evolutionary calculations work.

---

<div class="post-metadata">

**Author:** ![Rumraket](https://sea2.discourse-cdn.com/flex016/user_avatar/discourse.peacefulscience.org/rumraket/32/9328_2.png) [@Rumraket](https://discourse.peacefulscience.org/u/Rumraket)\
**Post date:** [September 1, 2019, 4:45pm UTC](https://discourse.peacefulscience.org/t/comments-on-gpuccio-functional-information-methodology/7560/381 "2019-09-01T16:45:16Z")

</div>

> [@colewd](#):
>
> This problem is so large when you take into account all the machines in the eukaryotic cell that require coordinated function the implausibility by mutation and selection is an accurate description.
> 
> You can always imagine a way it could happen but if we look at what the evidence is showing us there is almost no way to make evolutionary calculations work.

Show the calculation then.

---

<div class="post-metadata">

**Author:** ![colewd](https://avatars.discourse-cdn.com/v4/letter/c/ecd19e/32.png) [@colewd](https://discourse.peacefulscience.org/u/colewd)\
**Post date:** [September 1, 2019, 4:50pm UTC](https://discourse.peacefulscience.org/t/comments-on-gpuccio-functional-information-methodology/7560/382 "2019-09-01T16:50:57Z")

</div>

> [@Rumraket](#):
>
> Show the calculation then.

50% substitutability for prp8 is 2335 bits since 10 different amino acids work in every position.

Try to find a condition where prp8 can evolve given human and mouse are 99% similar. You will see that preservation is a very relevant observation.

The function to sequence space window for evolution to work is minuscule. If you can find it now for prp8 you need the coordinated binding and function of another 200 proteins.

---

<div class="post-metadata">

**Author:** ![Rumraket](https://sea2.discourse-cdn.com/flex016/user_avatar/discourse.peacefulscience.org/rumraket/32/9328_2.png) [@Rumraket](https://discourse.peacefulscience.org/u/Rumraket)\
**Post date:** [September 1, 2019, 5:12pm UTC](https://discourse.peacefulscience.org/t/comments-on-gpuccio-functional-information-methodology/7560/383 "2019-09-01T17:12:07Z")

</div>

> [@colewd](#):
>
> 50% substitutability for prp8 is 2335 bits since 10 different amino acids work in every position.

Why is that a problem?

---

<div class="post-metadata">

**Author:** ![colewd](https://avatars.discourse-cdn.com/v4/letter/c/ecd19e/32.png) [@colewd](https://discourse.peacefulscience.org/u/colewd)\
**Post date:** [September 1, 2019, 5:19pm UTC](https://discourse.peacefulscience.org/t/comments-on-gpuccio-functional-information-methodology/7560/384 "2019-09-01T17:19:39Z")

</div>

> [@Rumraket](#):
>
> Why is that a problem?

The problem is that there is a minuscule window for evolution to work and gpuccio preservation data is showing that window is exceedingly unlikely. When the dust settles you will realize the only known generator of 2335 functional linear bits is a mind.

Out for the day. Take care.

[Previous page](https://discourse.peacefulscience.org/t/comments-on-gpuccio-functional-information-methodology/7560.md?page=15)

[Next page](https://discourse.peacefulscience.org/t/comments-on-gpuccio-functional-information-methodology/7560.md?page=17)