Sunday, February 17, 2013

How to Kill Evolution using DNA

One necessary requirement for proving a specific line of Common Descent
 among species is the presumed impossibility of 'mixture' across species.

Without this constraint, DNA could then hop across populatons
from species to species freely, and any gene or piece of code
could come from anywhere.   Tracing DNA would be impossible.

To test whether DNA could hop between species is actually very simple.

For our purposes, we'll call this the Transitivity Test.   If three species fail the Transitivity Test, no specific line of Common Descent is possible,
and some other explanation must be posited for the results of the test.

An example will make this clear:

We examine a common strip of DNA shared by three species.
Much of the strip is both identical and shared by all three,
so we are reasonably sure we are looking at the same zone:

                     (1)       (2)          (3)         (4)
Species A: .. CT.---  TGGA --- CAT --- CTGAGCCCC...
Species B: ...GT---  TGGA --- CGT --- CTGAGCCCC...
Species C: ...GT---  CGGG --- CAT --- CTGAGCCCC...

Suppose we want to test a specific possible Descent Tree:

We propose that Species C is the common ancestor of A and B:

The problem is,  each species shares some DNA with the others:

But according to our primary axiom,
that Species cannot exchange DNA directly,
this is already an impossible situation.
In fact, it is impossible, no matter which species
we make the 'common ancestor'.

One might say 'no problem; these three species
are not direct descendants of one another.

But unfortunately, it does not end here:
Not only is it impossible for these species to be directly descended,
it also requires that we actually explain how they came to share DNA,
if they were NOT descended from one another.

There can only be two basic possibilities:

(1)  They all inherited their DNA from a single common ancestor,
and the unique shared 'readings' indicate mutation (corruption of DNA), or,

(2)  By coincidence two species happened to have their DNA
mutated or corrupted at the very same spot and in exactly the same way!

Neither of these cases bode well for genetic methods of tracing genealogy:

If (1) is allowed, then no tree can be constructed with any certainty
in regard to descent, since we already can see that the DNA data is
too corrupted to use to construct genealogies.
The frequency or commonness of a reading is no longer any indication
of "when" it originated or even "if" a given reading is original or a corruption.

If (2) is allowed, we have to recognize something even worse;
We might explain a single corruption of one Nucleotide (one letter)
by mere bad luck:  Once a mistake occurs in the same spot,
the mistake has a 1 in 4 chance of matching a mistake in the other species.  But to match in exactly the same place is a whole order of magnitude more unlikely.  In real cases, we are going to find thousands of such 'mistakes',
stretching for much longer sequences than a single Letter in the DNA code.
These kinds of 'coincidences' become so unlikely as to make them
practically impossible.

In any case, it is preferable to uphold the hypothesis of a barrier
preventing DNA transfer between species generally, since it is
an observable trend, whereas coincidental mutations are statistically
highly implausible. 

But the cost is twofold: 

(1) no direct tree of descent will be allowable.

(2) A hypothetical 'common ancestor' (now lost) must always be added.

But this is not any kind of evidence that a common ancestor even existed.
Instead, many other possibilities are still left wide open, such as common building blocks, mechanical constraints, or a shared design.

With the less likely option, i.e.,  (2) Cross-species DNA exchange,
we have to contrive a new mechanism instead of a new ancestor.

In order to account for significant correspondences between two species,
not shared by other species, we can postulate another mechanism,
that would allow the transmission of DNA between species, defeating
speciation and its premise entirely.

The cost here is as follows:

(1)  The acceptance of a hypothetical transfer mechanism between species.

(2)  The abandonment of the axiom that DNA is not transferred between species.

In practice it is impossible to determine which case explains any given
set of matches.  Even if it were, the very fact of "mixture" makes
the reconstruction of specific genealogical trees of descent impossible.

But if NO case of genetic inheritance can be traced to descent,
then there can be no evidence of Evolution and descent from DNA analysis.

No comments: