Fallacy of the Phylogenetic Signal? Part 2

T_aquaticus · August 28, 2020, 3:42pm

This might be a good place to start:

Summary

1. Phylogenetic signal is the tendency of related species to resemble each other more than species drawn at random from the same tree. This pattern is of considerable interest in a range of ecological and evolutionary research areas, and various indices have been proposed for quantifying it. Unfortunately, these indices often lead to contrasting results, and guidelines for choosing the most appropriate index are lacking.

2. Here, we compare the performance of four commonly used indices using simulated data. Data were generated with numerical simulations of trait evolution along phylogenetic trees under a variety of evolutionary models. We investigated the sensitivity of the approaches to the size of phylogenies, the resolution of tree structure and the availability of branch length information, examining both the response of the selected indices and the power of the associated statistical tests.

3. We found that under a Brownian motion (BM) model of trait evolution, Abouheif’s C mean and Pagel’s λ performed well and substantially better than Moran’s I and Blomberg’s K . Pagel’s λ provided a reliable effect size measure and performed better for discriminating between more complex models of trait evolution, but was computationally more demanding than Abouheif’s C mean. Blomberg’s K was most suitable to capture the effects of changing evolutionary rates in simulation experiments.

4. Interestingly, sample size influenced not only the uncertainty but also the expected values of most indices, while polytomies and missing branch length information had only negligible impacts.

5. We propose guidelines for choosing among indices, depending on (a) their sensitivity to true underlying patterns of phylogenetic signal, (b) whether a test or a quantitative measure is required and (c) their sensitivities to different topologies of phylogenies.

6. These guidelines aim to better assess phylogenetic signal and distinguish it from random trait distributions. They were developed under the assumption of BM, and additional simulations with more complex trait evolution models show that they are to a certain degree generalizable. They are particularly useful in comparative analyses, when requiring a proxy for niche similarity, and in conservation studies that explore phylogenetic loss associated with extinction risks of specific clades.
https://besjournals.onlinelibrary.wiley.com/doi/10.1111/j.2041-210X.2012.00196.x

EricMH · August 28, 2020, 4:50pm

this looks promising

from what i gather, the generate trees, walk traits along the trees with bifurcating brownian motion, and collect traits at the leaves. Then they interpolate the BM leaves with random distribution of traits on the leaves.

The null hypothesis is the tree metric measured on the leaves’ traits randomly permuted a bunch of times, and then the signal is measured as deviation from the null hypothesis distribution.

This should deal with my random leaves scenario, although a variable leaf length can maybe cause problems, and I will check what happens with the DAG.

Thanks, this is a big help!

per your request for a real design database, that would be ideal, but is beyond my time constraints

the best i can do is change genes to have more human inventiony sounding labels, and impose my preconceptions about real constraints on human designs, so will see what i can do there

EricMH · August 28, 2020, 6:08pm

also the phrasing should be changed to “decently nested clades” since no real world dataset exhibits a perfect tree structure

the “perfect nesting” is just an attribute of the algorithm used, and not an attribute of the dataset itself

this is what initially threw me, since it all sounds so perfect, and a perfect tree is certainly unexpected with intelligent design

but it is all just rhetoric, there is nothing perfect about the clade nesting

otoh an imperfectly nested clade is exactly what we expect from intelligent design, whereas from first principles we expect a perfectly or at least nearly perfect nested clade from evolution

so the actual imperfectly nested clade we see in the genetoc data may in fact be better evidemce for intelligent design than evolution

has anyone done such an analysis in mainstream phylogenetics? i bet not!

in which case my point still stands, the actual imperfectly nested clade we see in the data cannot be claimed as evidence of evolution without further analysis

T_aquaticus · August 28, 2020, 8:46pm

I like to describe it as a phylogenetic signal that sticks out above the noise caused by various mechanisms such as incomplete lineage sorting and homoplasies.

WHY??? Separately created species don’t need to share any features. If they do share features, there is certainly no expectation that those features would fall into a nested hierarchy. You would just as likely create a species with fur and flow through lungs as you would a species with feathers and teats. A nested hierarchy for intelligently design species makes no sense.

Chris_Falter · August 28, 2020, 9:09pm

Could you explain this in greater detail?

The prior of nested clades in the deep past is, AFAICT, based on extremely well-documented and evidenced observations of nested clades in recent geological eras.

In other words, biologists infer nested clades in the deep past in the same way and for the same reasons that physicists infer quantum dynamics in the deep past.

Peace,
Chris

EricMH · August 29, 2020, 2:52am

In the same way separate inventions share same features, there are functions that are common across widely disparate domains. E.g. both a thermometer and a car nowadays are going to have a microchip. Common functional needs entail common designs.

Bill_II · August 29, 2020, 9:58pm

Common functional needs don’t require common designs. Humans do it for efficiency or cost savings. Reasons that wouldn’t apply to an Intelligent Designer that is capable of creating life.

EricMH · August 30, 2020, 1:43am

even an omnipotent intelligent designer is constrained by logical necessity, and is self constrained by consistency

EricMH · August 30, 2020, 1:50am

good news everyone, i’ve figured out how to derive DAGs at the sequence level

soon all our questions shall finally be answered for ever

stay tuned

Klax · August 30, 2020, 7:31am

We’ll have economic nuclear fusion in 25 years too.

Bill_II · August 31, 2020, 3:01pm

The logical is based on human understanding and yet the intelligent designer is omnipotent? That does not compute.

Again, the ID is omnipotent so why should consistency be required? We’re not talking about squaring a circle here. The ID, if it exists, should be capable of just about anything. Like creating male mammals that are not subject to breast cancer.

T_aquaticus · August 31, 2020, 3:04pm

Why would an all powerful and all powerful deity need to reuse a single design? The only reason humans reuse designs is because we have limited knowledge and limited resources, none of which are limits on the supposed designer.

T_aquaticus · August 31, 2020, 3:06pm

Huh??? That sounds like babble.

Why does the tRNA for methionine need to have a UAC anticodon that matches an ATG codon? Why couldn’t it be some other codon?

On the flip side, why design millions of different cytochrome c all with different sequences?

EricMH · August 31, 2020, 3:29pm

sure an omnipotent designer can whatever

but intelligent design isn’t interested in that question

it is more about what we see in the genetic data

T_aquaticus · August 31, 2020, 3:32pm

EricMH · August 31, 2020, 3:36pm

if it is true, doesn’t matter what it is used for

main question is whether the ideas have scientific validity

they seem to afaik

T_aquaticus · August 31, 2020, 3:48pm

If it is said by design manifested in special creation, we must suppose that the Deity formed an archetypal plan of certain structures, and that He determined to adhere to this plan through all the modifications which those structures exhibit. Now the difficulties in the way of this supposition are prodigious, if not quite insurmountable. In the first place, why is it that some structures are selected as typical and not others? Why should the vertebral skeleton, for instance, be tortured into every conceivable variety of modification in order to make it serviceable for as great a variety of functions; while another structure, such as the eye, is made in different sub-kingdoms on fundamentally different plans, notwithstanding that it has throughout to perform the same function? Will any one have the hardihood to assert that in the case of the skeleton the Deity has endeavoured to show His ingenuity by the manifold functions to which He has made the same structure subservient; while in the case of the eye He has endeavoured to show his resources by the manifold structures which He has to subserve the same function? If so, it appears to me a most unfortunate circumstance, that throughout both the vegetable and animal kingdoms, all cases which can be pointed to as showing ingenious adaptation of the same typical structure to the performance of widely different functions, are cases which come within the limits of the same natural group of plants and animals, and therefore admit of being equally well explained by descent from a common ancestry; while all cases of widely different structures performing the same function are to be found in different groups of plants or animals, and are therefore suggestive of independent variations arising in the different lines of hereditary descent.
–George Romanes, “Scientific Evidences of Organic Evolution”, 1882
The Project Gutenberg eBook of The Scientific Evidences of Organic Evolution, by George J. Romanes, M.A., LL.D., F.R.S.

This is a 100+ year old problem that ID has yet to solve.

EricMH · August 31, 2020, 3:58pm

ID approaches it from the opposite direction, not making assumptions about how a god did things, or even if a god did things

ID could theoretically be due to aliens or something

but you know all this already

T_aquaticus · August 31, 2020, 4:02pm

You are assuming the designer had to reuse designs.

EricMH · August 31, 2020, 4:02pm

some obviously do, e.g. i import libraries all the time writing code

designer doesn’t have to work this way, but it is certainly consistent with what we know about intelligent design