Early patterns of the TB.

Understanding pedigrees, inbreeding, dosage, etc.

Moderators: Roguelet, hpkingjr, WaveMaster, Lucy

User avatar
Pan Zareta
Breeder's Cup Winner
Posts: 2074
Joined: Wed Dec 22, 2004 10:55 am
Location: west TX boonies

Postby Pan Zareta » Mon Dec 08, 2008 10:59 pm

vineyridge wrote:The reason that I asked about 5-e and 10-e is that both have a good history of mares who pass on very high quality jumping talent through generation after generation. If 5-e is a family that descends from Northern European mares rather than Oriental mares, I was sort of wondering if the same were true of 10-e. One sample from the whole family isn't really enough to make any call from, though.


Although the map (Fig. 2, p.292, Hill et al. in Animal Genetics, 2002) was somewhat suggestive of a geographic association between Northern European stock and haplotype L (found in all samples from fam. 5 exclusive of desc. of Hag 1744), Jansen's findings (PNAS 99:16 10905-10910, 2002) indicate greater geographic diversity in that area of the phylogenetic network. Haplotype M (found in desc. of Hag, and identical betw bp 15476-15818 to the 2008 submission defined as TB-5e) was not mapped, but would be nearer A than L.

Ref. GenBank sequences associated w/ the aforementioned study by Jansen et al., the TB is not alone in conserving Hill's haplotype F in two forms, one w/ a deletion betw bp 15528-32. By ignoring all indels the reports by Hill et al. 2002 and Harrison/Turrion-Gomez 2006 were faithful to the phylogenetic model of evaluating non-recombinant DNA, but that model is a pretty blunt instrument with which to analyze the matrilineages of a breed only ~350 yrs. old. Personally, I think that by ignoring the deletion a major point of segregation was overlooked. It's quite plausible that this is responsible for the superficially singular finding of diversity in functional mtDNA among groups "identical" by d-loop reported by Harrison/Turrion-Gomez.

Additionally, the relatively few GenBank accessions of longer d-loop sequences, for example those submitted by Flannery & Cothran, UKy 2006 (unassociated w/ any published report) show numerous segregating sites above bp 15827 (the high end of the sequence used by Hill et al. to analyze the TB haplotypes). None of these are meaningful from the broader phylogenetic viewpoint, but a handful of them look like they might be helpful in distinguishing betw TB families, imho.

aethervox
Allowance Winner
Posts: 395
Joined: Sat Mar 15, 2008 11:39 am

Postby aethervox » Tue Dec 09, 2008 3:24 pm

On a totally different topic, I've discovered a little bit about the 2008 study of thoroughbred dam lines. It appears to be part of a larger study that's explained at http://www.mcdonald.cam.ac.uk/projects/genetics/mimbower/mimbower.htm - Genetic variation in historic elite Thoroughbred horses.

As a part of the study they're doing DNA testing on Eclipse and Hyperion! :shock:

aethervox
Allowance Winner
Posts: 395
Joined: Sat Mar 15, 2008 11:39 am

Postby aethervox » Tue Dec 09, 2008 9:48 pm

After re-downloading the Hill sequences, I decided to compare them to Jansen's Clades (Table 2 in the PNAS paper referenced by Pan Zareta which can be found at http://www.pnas.org/content/99/16/10905.abstract?sid=70c1a5a4-8016-408d-8ffa-99aa87253445)

Here's what I came up with - no match means that the sequence didn't match Jansen's Clades or Hill's Haplotypes, depending upon which column they are in. The sequences are sorted by Family.

Sequence * Fam * Jansen Clade * Hill Haplotype
AF481305 * 1 * C2 * H
EU580148 * 1 * C2 * H
AF481306 * 2 * D3 * F
EU580149 * 2-8-16 * D3 * F
AF481307 * 3 * D2 * E
EU580152 * 3-15 * D2 * E
EU580153 * 3-18 * A5 * No Match
AF481308 * 4 * B2 * J
EU580154 * 4-11-13 * B2 * J/P
AF481309 * 5 * No Match * L
EU580156 * 5 * No Match * L
EU580155 * 5-e * No Match * M
AF481310 * 6 * D2 * C
EU580158 * 6 * D2 * C
EU580159 * 6-20-23 * C1 * N
AF481311 * 7 * D3 * F
EU580160 * 7-17-22 * D3 * F
AF481312 * 8 * D3 * F
EU580161 * 8c * No Match * No Match
AF481313 * 9 * A1 * A
EU580162 * 9 * A1 * A
EU580151 * 9-12 * No Match * D
AF481314 * 10 * D2 * B
EU580163 * 10-14-42 * D2 * B
AF481315 * 11 * B2 * J
EU580164 * 11f * B2 * J/P
AF481316 * 12 * No Match * G
EU580165 * 12b * A4 * No Match
AF481317 * 13 * B2 * J
AF481318 * 14 * D2 * D
AF481319 * 16 * D3 * F
AF481320 * 17 * D3 * F
AF481321 * 19 * No Match * K
EU580157 * 19 * No Match * K
EU580150 * 20 * No Match * O
EU580166 * 21 * No Match * No Match
AF481322 * 22 * D3 * F
EU580167 * 23 * B2 * No Match
AF481323 * 25 * C2 * I
EU580168 * 25 * C2 * I
EU580169 * 26 * No Match * No Match
EU580170 * A4 * C2 * H
EU580171 * B3 * D3 * F
EU580172 * B4 * D3 * No Match

More info as I have time to play with sequences and numbers :)

Edited (edits in red) to reflect corrected Clades - error pointed out by Pan Zareta :oops:
Last edited by aethervox on Wed Dec 10, 2008 9:04 am, edited 1 time in total.

User avatar
Pan Zareta
Breeder's Cup Winner
Posts: 2074
Joined: Wed Dec 22, 2004 10:55 am
Location: west TX boonies

Postby Pan Zareta » Wed Dec 10, 2008 12:06 am

aethervox wrote:EU580159 * 6-20-23 * No Match * N


By sequence motif, hap. N belongs in Jansen's cluster C1.

aethervox
Allowance Winner
Posts: 395
Joined: Sat Mar 15, 2008 11:39 am

Postby aethervox » Wed Dec 10, 2008 9:06 am

Pan Zareta wrote:
aethervox wrote:EU580159 * 6-20-23 * No Match * N


By sequence motif, hap. N belongs in Jansen's cluster C1.


You are correct. :oops:

I've edited the table to correct the errors. I was one cell off in Excel.

Thanks for pointing it out.

User avatar
Pan Zareta
Breeder's Cup Winner
Posts: 2074
Joined: Wed Dec 22, 2004 10:55 am
Location: west TX boonies

Postby Pan Zareta » Wed Dec 10, 2008 2:57 pm

aethervox wrote:I was one cell off in Excel.


Been there. Done that. :wink:

It caught my eye b/c my hunter mare is fam. 6a, Cream Cheeks' line, the branch in which hap. N was identified, 2002, so I'd already checked to see where she might be in Jansen's clusters.

Have you compared the sequences for accessions AF481310, 2002 defined as TBF6, and EU580158, 2008 defined as TBG-6? TBF6 has an insertion at bp 15598, and varies from the ref. seq., as well as from TBG-6, at 15578 and 15592. No apparent reason presents itself (at least not to me) for the omission of the variable bp's from Table 2., p. 290, of the 2002 report. I'm wondering if a sequence was submitted under the wrong definition, but at all other positions betw 15476 and 15818 AF481310 is a match for the sequence as given for haplotype C, Table 2.

aethervox
Allowance Winner
Posts: 395
Joined: Sat Mar 15, 2008 11:39 am

Postby aethervox » Wed Dec 10, 2008 3:42 pm

Pan Zareta wrote:Been there. Done that. :wink:

It caught my eye b/c my hunter mare is fam. 6a, Cream Cheeks' line, the branch in which hap. N was identified, 2002, so I'd already checked to see where she might be in Jansen's clusters.

Have you compared the sequences for accessions AF481310, 2002 defined as TBF6, and EU580158, 2008 defined as TBG-6? TBF6 has an insertion at bp 15598, and varies from the ref. seq., as well as from TBG-6, at 15578 and 15592. No apparent reason presents itself (at least not to me) for the omission of the variable bp's from Table 2., p. 290, of the 2002 report. I'm wondering if a sequence was submitted under the wrong definition, but at all other positions betw 15476 and 15818 AF481310 is a match for the sequence as given for haplotype C, Table 2.


Well, Hill does say that they ignored indels. After asking one of the systematics people where I work about indels, I was told that there were no set rules for when to ignore indels, etc. and that, depending upon what slant the writer wanted to take, they could emphasize or de-emphasize certain indels.

I suspect that they didn't use those basepairs in the 2002 table because they didn't need those sites to differentiate that haplotype, but that's just a guess on my part. It will be interesting to see if EU580158 shows up as a different haplotype in the unpublished paper.

One thing that caught my attention is that there appear to be two versions of haplotype F - one with a gap at 15528 and one without. The non-gapped version shows up in families 2, 8, and 16, and the other shows up in families 7, 17 and 22. The names given to EU580149 (2-8-16) and EU580160 (7-17-22) seem to indicate that they may now consider the indel at 15528 important. Only time (and publication of the paper) will tell. :D

User avatar
Pan Zareta
Breeder's Cup Winner
Posts: 2074
Joined: Wed Dec 22, 2004 10:55 am
Location: west TX boonies

Postby Pan Zareta » Wed Dec 10, 2008 7:33 pm

aethervox wrote: After asking one of the systematics people where I work about indels, I was told that there were no set rules for when to ignore indels, etc. and that, depending upon what slant the writer wanted to take, they could emphasize or de-emphasize certain indels


Even though Hill et al. 2002 ignored indels for making inter- and intra-breed comparison, their indel-containing sequences were submitted 'unedited', except for length, to GenBank. Is that usually the case, or might other equine d-loop sequences therein, submitted by other researchers, have been 'purged' of indels before submission?

One thing that caught my attention is that there appear to be two versions of haplotype F - one with a gap at 15528 and one without. The non-gapped version shows up in families 2, 8, and 16, and the other shows up in families 7, 17 and 22. The names given to EU580149 (2-8-16) and EU580160 (7-17-22) seem to indicate that they may now consider the indel at 15528 important. Only time (and publication of the paper) will tell. :D


The TB is not alone in conserving 'dual versions' of haplotype F. Imho, that deletion has probably been conserved a very long time, maybe even longer than the avg. mutation rate in 'functional' equine mtDNA.

aethervox
Allowance Winner
Posts: 395
Joined: Sat Mar 15, 2008 11:39 am

Postby aethervox » Wed Dec 10, 2008 8:17 pm

Pan Zareta wrote:Even though Hill et al. 2002 ignored indels for making inter- and intra-breed comparison, their indel-containing sequences were submitted 'unedited', except for length, to GenBank. Is that usually the case, or might other equine d-loop sequences therein, submitted by other researchers, have been 'purged' of indels before submission?


I believe they are unedited; they can be updated if later research shows the original sequence to be in error.

The TB is not alone in conserving 'dual versions' of haplotype F. Imho, that deletion has probably been conserved a very long time, maybe even longer than the avg. mutation rate in 'functional' equine mtDNA.


Does that haplotype occur in other equids, like donkeys or zebras? It'd be interesting to know if it's a horse-only haplotype.

User avatar
Pan Zareta
Breeder's Cup Winner
Posts: 2074
Joined: Wed Dec 22, 2004 10:55 am
Location: west TX boonies

Postby Pan Zareta » Thu Dec 11, 2008 10:34 pm

aethervox wrote:Does that haplotype occur in other equids, like donkeys or zebras? It'd be interesting to know if it's a horse-only haplotype.


I haven't looked at other equids. What I've been trying to do is get an idea of how much diversity the haplotypes identified by Hill et al. 2002 might encompass. Studying the longer (976bp) GenBank sequences submitted '06 by Cothran & Flannery suggest some of them might more accurately be termed haplogroups. For example, I looked at 5 sequences (AY246230 AY246217 AY246216 AY246193 AY246192) from 3 breeds (Friesian, Clydesdale, Belgian), identical to the sequence betw 15476-15818 for haplotype F (the version w/out deletion), and found several variable bp above that region effectively defining five different haplotypes w/in haplotype F, i.e. cluster D3.

xfactor fan
Breeder's Cup Winner
Posts: 2212
Joined: Thu Sep 16, 2004 8:46 pm

Postby xfactor fan » Fri Dec 12, 2008 10:17 am

Why did the two types of "F" get classified together? With a split that far back there are two clearly different linages. Shouldn't there be a F1 & F2?

aethervox
Allowance Winner
Posts: 395
Joined: Sat Mar 15, 2008 11:39 am

Postby aethervox » Fri Dec 12, 2008 10:54 am

xfactor fan wrote:Why did the two types of "F" get classified together? With a split that far back there are two clearly different linages. Shouldn't there be a F1 & F2?


A lot depends on what the researcher(s) are looking for. Since Hill, et. al. were only trying to see whether haplotypes were conserved within dam lines, it really didn't matter to them whether there were subgroups of haplotype F. With the Bower paper (2008 study with hill) it will be interesting to see whether that changes.

User avatar
Pan Zareta
Breeder's Cup Winner
Posts: 2074
Joined: Wed Dec 22, 2004 10:55 am
Location: west TX boonies

Postby Pan Zareta » Fri Dec 12, 2008 11:44 pm

aethervox wrote:
xfactor fan wrote:Why did the two types of "F" get classified together? With a split that far back there are two clearly different linages. Shouldn't there be a F1 & F2?


A lot depends on what the researcher(s) are looking for. Since Hill, et. al. were only trying to see whether haplotypes were conserved within dam lines, it really didn't matter to them whether there were subgroups of haplotype F. With the Bower paper (2008 study with hill) it will be interesting to see whether that changes.


Imho, the discussion section of the 2002 report reflects a high degree of confidence that all individuals with "haplotype" F are completely identical by d-loop, indels notwithstanding, and share a very recent common founder. ".....[W]e have uncovered a web of founder sharing" presumes as much, considering the fact that individuals who really are completely identical by d-loop mtDNA may still be many thousands of years distant from their most recent common ancestress.

User avatar
diomed
Grade III Winner
Posts: 1142
Joined: Mon Oct 04, 2004 4:16 pm

Postby diomed » Thu Jan 01, 2009 6:43 pm

Great discussion about the current MtDNA research results, etc..
Sometimes I am lost a bit, but am fascinated nonetheless. :lol:
Anyways, I have been doing a lot of research into the TB/QH foundations and would love to know more about Janus II and the clan of family 15.
I noticed that many founding QH mare lines are also HUGE TB families as well.
Most, if not all go back to a Janus mare.
Anyways, I wanted to get a hold of Barbara Muse's book on the QH breed.
I believe it was originally named "The Grand Twenty".
Unfortunately I cannot afford to buy any book right now and was wondering if anyone here has read it and can offer me info about it.
Is it really the best source on Janus II?

Till next time.....

User avatar
Pan Zareta
Breeder's Cup Winner
Posts: 2074
Joined: Wed Dec 22, 2004 10:55 am
Location: west TX boonies

Postby Pan Zareta » Thu Jan 01, 2009 8:53 pm

diomed wrote:Great discussion about the current MtDNA research results, etc..
Sometimes I am lost a bit, but am fascinated nonetheless. :lol:
Anyways, I have been doing a lot of research into the TB/QH foundations and would love to know more about Janus II and the clan of family 15.


Judging from GenBank sequences submitted 2008 by Bower et al. (incl. E.W. Hill, lead author of the 2002 study) one or more individuals from fam. 15 have been tested and found to have the same mtDNA haplotype at between bp 15476-15818 as Hill's haplotype 'E' (found in some members of fam. 3 and consistent w/ Jansen's cluster D1). That doesn't necessarily mean families 3 and 15 share a tail female ancestor in modern times (>1000 AD) or are identical at full mtDNA control region sequence, tho'.

Anyways, I wanted to get a hold of Barbara Muse's book on the QH breed.
I believe it was originally named "The Grand Twenty".

Correct.
Is it really the best source on Janus II?

:shock:
NO!

Great source for photos seldom seen elsewhere, but extremely unreliable as to content.

Imho, the best and most reliable source on Janus is Fairfax Harrison's Early American Turf Stock.

PZ