How Much of Genome is "Functional" or "Neutral"?

[Ashwin_s]

@Faizal_Ali

Is the whole chromosome then functional, according to your definition?

The gene is functional.
If the rest of the chromosome will not be functional because of the gene.

[Faizal_Ali]

@Ashwin_s

I think you didn’t understand Graur’s definition. If the feature is neutral (or useless); then as per Graur, it doesn’t have function.

This is getting tedious. I will repeat Graur’s definition one more time. Please try to read for comprehension.

A genomic segment is considered to possess a selected effect function if at least one out of all the possible mutations that can affect its sequence is deleterious.

How do you understand the phrase “at least one out of all the possible mutations”?

[Faizal_Ali]

@Ashwin_s

The gene is functional.
If the rest of the chromosome will not be functional because of the gene.

Same as Graur’s definition. So what are you complaining about?

[John_Harshman]

@Ashwin_s

Would you agree?

I agree that it demonstrates the difficulty of defining “functional”. But you have been no help in that regard.

@Ashwin_s

Whether a feature is useless or not depends on the environment.
In a causal sense, the gene is functional.

Define “functional” in a causal sense. Are the eyes of cave fish functional? And what about a feature that’s useless in all environments?

@Faizal_Ali

For instance, the allele that causes eyes to develop may be deleterious because it wastes biological resources on structures that produce no benefit. That would make the sequence (as defined by all its variants) functional, because a deleterious form exists.

That seems like a nonfunctional definition, then, because just about any sequence could have a deleterious form. Any patch of junk DNA could gain a mutation that might turn it into a transcription factor binding site that would adversely affect some metabolic or developmental pathway, for example. By that definition, most of the genome is indeed functional. What I think Graur may have meant to refer to is a sequence that’s maintained by purifying selection. Perhaps there is no hard and fast definition, and a functional sequence is one for which the density of possible deleterious mutations is high.

[swamidass (S. Joshua Swamidass)]

I think what is missed by IDists about ENCODE is that it reduces the difference between the “dark matter” of the genome and full-on-coding genes. If there is a low background transcription and translation of the genome, then (1) its low so it won’t affect things too much, but (2) at times, it might affect things enough to safely produce meaningful variation, and (3) gives a far more plausible pathway to the evolution of de novo genes.

Far from causing problems for evolutionary theory, the ENCODE data might provide resolution to several open questions.

[Ashwin_s]

@John_Harshman

I agree that it demonstrates the difficulty of defining “functional”. But you have been no help in that regard.

I am not a biologist… why would I be of help?
However, I am not buying a nonsensical/arbitrary definition. Causal roles look far more promising.

@John_Harshman

Are the eyes of cave fish functional? And what about a feature that’s useless in all environments?

The parts of the eye that can still play a causal role in sight are functional (even though the larger function of sight is lost due to failure of other parts). For example, in humans blindness might be caused by the lens of the eyes being damaged. This does not mean that the nerves that transmit information from the eyes to the brains are non-functional.

Can you give an example of a “feature” that’s useless in all environments?

[Ashwin_s]

@swamidass

Far from causing problems for evolutionary theory, the ENCODE data might provide resolution to several open questions.

Can you specify what questions you are referring to?

[glipsnort (Steve Schaffner)]

@Faizal_Ali

This seems to be demonstrating confusion associated with ambiguity of the term “sequence”. The sequences in question are presumably related to development of the brain. As such, if we stick with Graur’s defintion, there are almost certainly mutations in these sequences that will be deleterious. e.g. that would lead to the related structures of the brain being highly underdeveloped, or not developing at all.

Changes to some stretches of sequence will result in highly underdeveloped brains, while changes to other stretches will result only in slightly less efficient or less complex brains. In the kind of environment many of us now live in, those brains are selectively advantageous even though they function less well for thinking.

[Faizal_Ali]

@glipsnort

Changes to some stretches of sequence will result in highly underdeveloped brains, while changes to other stretches will result only in slightly less efficient or less complex brains. In the kind of environment many of us now live in, those brains are selectively advantageous even though they function less well for thinking.

So those meet Graur’s definition for function.

[swamidass (S. Joshua Swamidass)]

@Ashwin_s

Far from causing problems for evolutionary theory, the ENCODE data might provide resolution to several open questions.

Can you specify what questions you are referring to?

For example, where do de novo genes come from in higher order mammals?

If most the genome is engaged in low risk “experimentation” in transcription/translation (at too low a level to create real problems), then de novo genes might be much easier to stumble upon. When they are stumbled upon, it then merely becomes a task of amplifying and tuning expression, which is much easier to do.

Evolving a new gene used to be a very difficult process to conceive of. There are so many sequence components that are “required” for it: promoter, introns, exons, splice site, etc." However, in ENCODE we find that there is a large amount of the genome that seems to have no problem being transcribed and translated, albeit with incomplete sequence elements and also at a very low level. That makes the definition of a “gene” and their origins a far less a black and white binary problem, and more of a dimmer switch, with a far smoother fitness landscape.

This all works because there is (1) biochemical function and variation, that (2) is usually invisible to selection because it does not usually affect cellular function much, and (3) is only a hair’s step away from becoming a full fledged gene if it is helpful. So ENCODE, rather than challenging evolution, makes more sense out of it.

[Faizal_Ali]

@John_Harshman

That seems like a nonfunctional definition, then, because just about any sequence could have a deleterious form. Any patch of junk DNA could gain a mutation that might turn it into a transcription factor binding site that would adversely affect some metabolic or developmental pathway, for example.

That’s a good point.

[glipsnort (Steve Schaffner)]

@Faizal_Ali

So those meet Graur’s definition for function.

Of course. But under Graur’s definition intelligence-lowering mutations increase function, while for someone who is interested in the functioning of brains they decrease function. And mutations that make us nearsighted, which may well have no effect on fitness at all, have no effect on function by Graur’s definition, while for someone who is interested in the functioning of eyes they do indeed affect function.

[Giltil (Gilbert Thill)]

@glipsnort

Then why are you discussing ENCODE in connection to function at all? Whatever Dan Graur does or doesn’t think, ENCODE doesn’t tell us anything about function in the sense you’re talking about, right? And it doesn’t tell us anything about the plausibility of evolution, agreed? So what is the point of this exchange?

The connection with ENCODE is the following: in order for a segment of DNA to be fonctional (in the sense of contributing to fitness), it must have some sort of biochemical activity, which seems to be how ENCODE define function. So the discovery by ENCODE that at least 80% of the human genome has some sort of biochemical activity give more credit to the idea that a larger fraction of the human genome than traditionnaly estimated is functional ( in the sense of contributing to fitness)

[John_Harshman]

@Ashwin_s

Causal roles look far more promising.

Again, what do you mean by “causal roles”?

@Ashwin_s

The parts of the eye that can still play a causal role in sight are functional (even though the larger function of sight is lost due to failure of other parts).

So in order to determine functionality you have to take a feature apart into all its separate bits. How do you define “function”?

@Ashwin_s

Can you give an example of a “feature” that’s useless in all environments?

Sure. How about the damaged lens you just mentioned?

[pnelson (Paul A Nelson)]

Here’s the talk that I mentioned above (in response to John Harshman’s question):

https://www.discovery.org/f/41175/

[Giltil (Gilbert Thill)]

@swamidass

This all works because there is (1) biochemical function and variation, that (2) is usually invisible to selection because it does not usually affect cellular function much, and (3) is only a hair’s step away from becoming a full fledged gene if it is helpful. So ENCODE, rather than challenging evolution, makes more sense out of it.

Given what you are describing here, do you think that the human race is going to evolve toward a new species? And in that case, what will it mean to be human ?

[Giltil (Gilbert Thill)]

@pnelson

Here’s the talk that I mentioned above (in response to John Harshman’s question):

https://www.discovery.org/f/41175/

A lot of food for thought here. Thanks.

[swamidass (S. Joshua Swamidass)]

@Giltil

Given what you are describing here, do you think that the human race is going to evolve toward a new species? And in that case, what will it mean to be human ?

I note that your questions are far more insightful than your pronouncements.

These are good enough questions to pause and reflect before answering more.

[John_Harshman]

Could you perhaps summarize the relevant points from that presentation?

[Faizal_Ali]

@glipsnort

But under Graur’s definition intelligence-lowering mutations increase function, while for someone who is interested in the functioning of brains they decrease function.

Incorrect. If a sequence has a variant (alleles) that is deleterious, it is functional. That applies to all its alleles, whether they are neutral, beneficial, or deleterious.

You are making the same mistake @Ashwin_s is, of confusing “functional” and “beneficial”.