Rhiza Labs FluTracker Forum

maybe there was one ancestor for the Eurasian and American lineages
(~100 years ago)
and then it evolved separately in the blocks, while some particular
species or location did not mutate on the amino-acid-level

or there is a "valley" there, a PB2 which is superior to most others
and allows some evolution only inside the valley. Once it climbs
up the surrounding mountains, it falls back earlier or later.

Only in mammals then can it climb the mountains
(... and maybe in some centuries drop into another neighboring valley )

America and Eurasia are trapped in the same valley, but with different
starting versions of nucleotides.
These two versions emerged in the decades,centuries before.


well, the Altai duck and the Dutch vet were recent, only some years ago.
And still almost equal to the index and the number of nucleotide-differences
similar to the assumed ancient viruses near the index....

it's complicated ...

_________________
no patents on genes, publish the GISAID sequences !

The valley is named "atractor bassin", in the context of nonlinear dynamics.
That is exactly the picture I had in mind thinking about the avian index recurrence:

This is a simple one-dimensional picture. This is usually generalized to N dimensions.
In the genetics context, better use it as a metaphor.
My trouble with the index is that the valley should be too steep!
Something else "locks" some amino acid in an almost rigid way.

Yes, it is! There is a large number of bird species, and many different HA and NA.
The whole ecological system is SO COMPLEX that it is rather surprising to find
ANY definite pattern. The system has so many degrees of freedom!
However, some complex systems do show "self organizing patterns" (or synergy).
(Example: sometimes sky clouds form a regular patters ressembling "ribs')

That's the remarkable feature of the index approach: the pattern it is surprisingly simple.

we are still only talking about segment 1, no HA or NA.


Eurasian avian segment 1 separate into two groups , G and K
see here: http://www.flugenome.org/show_subtypes.php

they have
G:359
K:1222

I have excluded Australia and got G:291, K:725 (contains H5N1 etc.)

the 2nd cloud from top consists of pairs from these 2 groups
(G,K) or (K,G)

for pairs (K,K) and (G,G) see the attached graph

you can see, that both groups split again ...


both Eurasian groups ,G and K, although much different
on the nucleotide-level , do again both contain PB2s
which have amino-acid sequences identical to the index

_________________
no patents on genes, publish the GISAID sequences !

Is I understand, the isolates universe is now quite focused.
I have to check further the link and pairs definitions to get
better awareness and clearer insight into this new image.
(suggested name to pairwise pictures: Distance Distribution Diagram)

It is amazing how many features of the very first picture reappear.
The “second split” is an entirely new feature (2 clouds on the right side).
It is interesting to see the picture counterpart of different filters.

I have a question on the upper cloud of the first (unfiltered) plot:
http://magictour.free.fr/panflu/sepp1c0.JPG
All the pairs of this cloud are between North American lineages?
(no cross American x Eurasian). Is that correct?
(Cluster or cloud inside range 300<N<400 and 0<A<35)
This cluster disapears on Eurasian only picture.

no, in
http://magictour.free.fr/panflu/sepp1c0.JPG
the upper cloud is of pairs where one component is North-American(NA)
and the other Eurasian(EA). So : pairs (EA,NA) and (NA,EA)

The 2nd and 3rd pictures contain the pairs (NA,NA)
and (EA,EA)

EA splits into groups G (EAG) and K(EAK). EAK is bigger and
contains most of the H5N1-PB2s

The 2nd cloud is of pairs (EAG,EAK) or (EAK,EAG)

The new picture
download/file.php?id=145
is of pairs(EAK,EAK) on the left and of pairs(EAG,EAG) on the right.
The former upper clouds are now missing in that picture, of course.


remember, this is still all only segment 1.


time to consider time, which was ignored so far

_________________
no patents on genes, publish the GISAID sequences !

OK, It is clearer now! All of the pairs of this cloud are a continent appart.
None is very close at nucleotide level. THIS explains the 340 Nucleotides difference “offset”.
“Eppur NON si muove”, And yet so many are very close at amino acid level.
The avian “golden standard” index!

> remember, this is still all only segment 1.
Yes, this is understood.

> time to consider time, which was ignored so far
Many variables involved. Not obvious which is the best 2D representation.
(perhaps use the index as reference)
BTW the pairwise representation is still far from exausted. Very interesting patterns!

The "return to index" behavior is defiant to the impression that evolution=improvement.
It means that sometimes the evolution time runs "backwards" (reverting amino acid level mutations)
although the Nucleotide molecular clock moves forward!

I can only explain it as that the working viral possibilities
in segment 1 (and 2,3,5,7,8 ?) are very limited for birds.

If the virus evolves too much, then it is no longer competitive.
So this optimal index-configuration somehow forces
the evolution in one direction.

Not so in humans,swine,horses
Not so much in South America,some Henan H9N2,
They should get some wild-bird sequences from Africa,South America


is this all known ? There should have been papers about it

_________________
no patents on genes, publish the GISAID sequences !

I think it's better to make at least one mammalian diagram (swine?; human?) to be sure.


This is the paper (*), IMO

at least under this graphic representation. I think there are previous papers on low S/N.
But, unless you missed some major paper during your years-long research, this sounds like brand new results!

(*)This is online ideas development and criticism. Classical paperwork publishing is much longer!
Well, the true paper starts a clear title. (sometimes we change title dozens of times).
Then, the abstract! How to compress the whole idea into a couple of paragraphs.
(rewrite some 150 times)
Then, background an exhaustive bibliographic review
(seeking bibliography, some 25 new ideas emerge).
Then, Material and methods. (Yes, can’t skip that. No way. Got to exhumate forgotten codes and fluxograms).
Well, then, the results. Scales, impeccable graphic presentation. Re-make same picture hundred times.
To explain step by step everything that is all too obvious to you in detail.
Conclusions. It’s an iterative process, sometimes with slow convergence.
Then, submit and wait... When you finally forgot the whole thing and just started a new pathway,
the awaited answer comes. The paper is accepted, but referee #1 claims different pictures are necessary,
while referee #2 wants you to re-write and expand “materials and methods”.
Referee#3 was clement and jus corrected a couple of references and added some extra ones.
Take a deep breathe! It is good news, paper was accepted. Publishing is (sometimes)
Already scheduled, so you have to “play it again”. You’re almost there, can’t stop now.
Re submit in a timely manner, and when you see it printed, you’re free!

I think the net can potentially modify the research dynamics. Or complement classical publishing.

just put it on a webpage, in whatever format you consider suitable !

Or better : a wiki, so others can add stuff and edit it. (must be approved)

And a discussion forum.

I like most wikipedia-articles more than .pdf - papers


that process should be superior to classical publishing.
Wait 100 years and we'll see ...

_________________
no patents on genes, publish the GISAID sequences !

We are already probing some of these features.
The post numbering and individual post links would take better advantage of intrinsic html hypertext vocation.
This is true to all threads, not research only. Many threads restarts topics already discussed on older ones. Cross referencing should be made easier!


A tentative rationale for the index-like genetic preservation:
This narrow margin to amino acid changes could be linked to the compatibility with a wide variety of avian hosts.
Index whould be a sort of “inter-species password”.
Since avian flu reassorts very frequently (as you mentioned before) each gene must have
the constraint of “broad host range portability”. This could generate the index-like stability.

I assume that avian hosts evolution is much slower than viruses. Almostly static in the virus evolution time scale.

Shortly: index key fit <=> ability to maintain broad circulation across avian hosts variety. The key would the be on the hosts harboring the viruses, not on the viruses exclusively!

(No such a constraint among “dead end” mammalians!)