Maybe this has already been discussed and I missed it, but there is a publication listing the genetic relatedness of 194 fig cultivars at the USDA National Clonal Germplasm Repository at UC Davis. You may find it here:
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2860561/pdf/10709_2010_Article_9442.pdf

They use a collection of microsatellite markers scattered through the fig genome to divide their collection into 10 groups genetically. They do find that a number of their cultivars are identical at this level of discrimination. You can see all that in Figure 1 in the paper. They find that Ischia Green, Monstrueuse, Paradiso and DFIC 208 are all the same as are Violette de Bordeaux, Vista and Beers Black. Anyway, Figure 1 makes for some fun viewing and the paper is an interesting read.

There are similar reports about other fig germplasm banks some of you might enjoy too.
http://journal.ashspublications.org/content/133/1/69.full.pdf+html
http://hortsci.ashspublications.org/content/40/1/29.full.pdf
http://www.historiambiental.org/files/publicaciones/GG_analysis.pdf
http://onlinelibrary.wiley.com/doi/10.1111/j.2007.0018-0661.01967.x/full

Good luck with your trees!

The problem is how people interpret this.  The figs are not necessarily genetically identical, it's just that in the few places they look they're the same.  It becomes much more expensive to look in more places so they don't do it.  The differences they don't look at may be minor or crucially important.

They choose which places to look by which ones are the most different.  As an example, 5 places may account for 40% of the variation, 10 places may account for 65% of the variation.   To get to 80% of the variation accounted for you might have to look at 25 places.  They decide how much of the variation to account for by how much it costs to get that extra part covered.  Hope that makes a little sense.

Bob,

That does make sense and is in line with what I have been thinking about so many of our cultivars. Those that are closely related, or by the standards the tesers are using, identical, are likely all from the same area, say the same town and surrounding area in whatever country they came. from 

Bob - when you say "places" are you talking about geographic places and if so, any idea how wide, geographically speaking, of an area do they check?

Good information

Bob,

You got me mixed on your last post. Are you discussing geographic as in cities and villages or are you referring to loci and allele genes?

I mean places in the genome. Physical spots on the chromosome.  Only figs from Turkey or Turkmenistan seem to be very different genetically from other figs.  The rest seem to have traveled around a lot.

Also, these are mathematical models of reality.  Different ways of building these models produce different results even with the very limited data being used. 

There's no estimate given of the % of the genome being analyzed, but my guess is it's much less than 1/100,000th of the whole genome.  That's not as bad as it sounds since, (as a wild guess) 60-90% of the genome of the family Moraceae (Mulberries, figs, 1400 species) will be the same.  To be efficient you want to look at the genes that have the most differences among figs.  But 20 tiny segments of 26 very long chromosomes is very, very few.  And that's the most anyone is looking at.  The longest piece of DNA in these studies is less than 400 base pairs.  The smallest plant genome is over 100 million base pairs.  The biggest study looked at approximately 6,000 base pairs.  I can't find info on the size of the fig genome, but rice has 390,000,000 base pairs and I would guess the fig has more.

Bottom line is our understanding of fig genetics is very primitive.  We're more at the stage of Galileo than NASA or even Newton (Sir Isaac, but the fig reference ties in nicely  :).  Some of what we think we're seeing isn't real and we're definitely not seeing the whole picture.  OTOH we're on our way, we just have to keep going  :)

Bob,

That begins to answer my questions about genetic investigation of different cultivars. It has always seemed to me that the % of DNA sampled was incredibly small and wondered how they determine which few alleles that they will use.

We see from this that Beer's Black, Violette de Bordeaux and Vista are the "same". But in you actual experience Vista is superior to any of  the VdB variants that I have.

We see that Orphan, 278-128 and Deanna are the "same" but when tasted at USDA/UC Davis WEO orchard, Orphan stands apart.

There is NO way you will mistake Encanto Brown Turkey for Archipel - not even close, but they show as the "same".

On the other hand, Brunswick, Red Italian and Rattlesnake Island are very similar and the DNA agrees.

Sorry I think I got a couple people confused with my talk about fig home towns. I was trying to relate my thoughts to the dna findings which as Bob explained very well, is concerned with chromosome location of that very tiny substance, dna on the genome of figs. 

I will try to clear-up the points being raised here. For what it is worth, molecular genetics and population genetics in oncology and immunology are a major part of my everyday professional life.

Bob, on one level you are correct. These results do not demonstrate that the cultivars showing identity by these tests are completely identical throughout their genome. You are also correct that they are focusing on areas of significant variation to define relatedness and not sequencing entire genomes. One aspect of that is cost (money, man power and time) as you point out. But another reason, and I'd say the primary one, is to as efficiently as possible define genetic relatedness in a scientifically rigorous way. I believe the group at Davis has done that in their publication.
Most of the genes from fig cultivar to fig cultivar vary very little. So, to examine relatedness, it is useful to find loci (specific places in a genome) where there is significant variability. Microsatellite DNA sequences (also known as VNTRs) fit this bill. What has been demonstrated repeatedly in both mammalian and plant systems, is that the amount of sequence variation associated with a microsatellite allele found on a particular chromosome is orders of magnitude (factors of 10) less than the variation in sequence associated with a different allele on the same chromosome. Let me try to give an example. Lets say the Chromosome 1 I got from my mother has a microsatellite locus that has 7 variants in the human population. Lets call the microsatellite variant I got from my mom 1A. If you DNA sequenced the entire genome from 100 other people carrying the 1A allele on their chromosome 1, you would find from to tens to maybe a couple of thousand base pair differences in the >103 million base pairs on Chromosome 1. A very small percentage (<10%) of that chromosome is DNA that codes for 3141 structural genes found on chromosome 1. So, the likelihood that there are meaningful phenotypic differences between people that carry the 1A allele is very low. When you then sequence the DNA of chromosome 1 from people who carry different alleles (versions) of the microsatellite loci (place on the chromosome), what you find is a MUCH higher rate of variation between allele 1A and 1B or 1C or 1D on to 1G, but essentially the SAME amount of variation among individuals that are 1B or 1C or 1D, etc. So, you would see tens of thousands to hundreds of thousands variations between the sequences from a 1A chromosome versus a 1B or 1C or 1D etc. If you pick loci where the variants (alleles) are broadly distributed within fig populations and any individual fig cultivar is likely to be heterozygous at that loci (that is have 2 different alleles), then you have a spot that measures diversity within the population of figs. If you next find a series of those alleles that behave independently relative to each other, you will hopefully generate a set that covers all 13 ficus carica chromosomes. There are 26 chromosomes in diploid cells. This is what the group at Davis has done. The math they use is strongly supported by experimental data. I am unaware of "different ways of building these models" that would "produce different results." To be sure, there has been refinement in the mathematics used to determine genomic relatedness over the past 30 years. But those refinements simply give better accuracy in determining relatedness and do not measure a new or different underlying phenomena. Based on my experience in looking at complete chromosome sequences from microsatellite loci identical individuals and microsatellite loci different individuals in human & mouse systems, my expectation is that, if you were to sequence 100 different Vista plants and 100 different Violette de Bordeaux plants from around the world or even just the US, the amount of variation between any 2 Vistas or any 2 VdBs would be no different than between any Vista and any VdB. On that basis it is hard to differentiate a Vista from a VdB. The same comparison would have no trouble differentiating a VdB or Vista from the closely related Negro Largo for instance.

I think the work with the microsatellite alleles does give a very clear view of cultivar relatedness at the genomic level. I agree there is a long way to go in defining individual genes in figs and how they function. And I got a real kick out of your historical analogies! I am still cracking up over (fig) Newton!!! I agree we are on our way and we just have to keep going.

Jon, I hope I've given some answer to your first statement. With respect to the "sameness" between Beer's Black, VdB and Vista, let me first say that I am very, very happy that I got a Vista from you instead of a VdB! It is certainly possible that there are minor genomic differences between Orphan, 278-128 and Deanna that result in Orphan producing a superior fig to your taste. However, you can also hypothesize explanations other than genetic differences. For instance, differences in watering or fertilizing or sun exposure might result in the difference you observe. Orphan could have a less damaging case of FMD from the Deanna and 278-128 you tasted. The process of going from tree genetics to how any individual rates the fruit is way complex with tons of factors other than genetic ones that might drive the outcome. Or it could all be genetic. Fodder for endless debate.

Good luck with your trees!

DWD2 re: Orphan, etc.

The orchard at WEO provides about as identical and environment as could be found. Same soil, climate, irrigation, fertilizer, etc as is humanly possible - all 300+ trees on 2 acres of flat ground. Not saying there aren't differences, only that they have been greatly minimized to the extent possible short of a hideously expensive controlled-environment study of some sort.

The whole subject is still a work in progress.

Thanks. Point taken. It is a reflex with me to try & generate alternate hypotheses.

DWD2 - great post!  I agree completely about efficiency.  But as you know efficiency comes at a price.  The difference between VdB and vista may come down to 1 less efficient version of an enzyme in a region that's generally well conserved.  Efficient methods will never find it, but there is (in that example) a genetic difference.  As for mathematical models, the UCD paper, for instance, talks about generating the 10 groups from cluster analysis.  They then state that if they analyze without cluster analysis they get different groups - I don't have the paper in front of me and I just got notice that my landlord sold my house and I have to move sooner than I expected so I must be brief   :)  My expectation is that with more loci used, you will find differences between Vista and VdB that aren't found with the 20 UCD used.  I believe 50 are now available.

Another monkey wrench is that histones, a protein coat around the chromosome, can control gene expression for several generations.  At least that's the case in people.  Is it the same in plants?

Bob - Sorry to hear your house got sold out from under you. Moving bites, pure & simple.

I think the chance that a minor change in an enzyme's activity would generate the differences people report between the cultivars shown to be synonymous by these methods is vanishingly small. Assuming the differences people report between cultivars that are identical by these methods are the result of genetic factors, I would guess that what you suggested in your last question is a much more likely reason. That is epigenetic causes, which are cases where two animals or two plants have very different patterns of expression of genes on at least one chromosome despite having identical DNA sequences. This can happen because of different histone modifications as you note, methylation of DNA sequences in a portion of one chromosome & not the other, methyl-CpG-binding proteins, transcriptional complexes and nucleosome remolding complexes just to name most of the big players. Epigenetic effects can cause the silencing of gene(s) on a stretch of a chromosome that are normal expressed or the expression of gene(s) along a stretch of chromosome that are normally silent in fruit development. 

I strongly doubt that using a larger array of microsatellite probes would provide meaningfully more discrimination. They do not say it in this publication but the probes were chosen from a library of >50 cloned and completely DNA sequenced microsatellite loci also referred to as simple sequence repeats (SSRs). The criteria for using a SSR to do genomic comparison of fig cultivars was that did not amplify multiple loci, they showed heterozygosity, and they assorted differently from other SSRs which meant the 2 SSRs are either on different chromosomes or far enough apart on the same chromosome that there was chromosomal recombination between the 2 loci during evolution allowing them to provide independent information. The point is, after checking, the remaining >30 loci apparently do not readily provide additional useful information.

The great part of discussions like this is they can provide the basis for future experiments. Lets assume that Orphan & Deanna do turn out to be identical by DNA sequence. If they taste very different when grown at WEO at Davis, you are then in a position to compare gene expression during fruit development to determine critical players in producing superior tasting fruit.

I thought that to include enough of the remaining loci to produce a significant increase was deemed too expensive.  I have a vague and unreliable memory of the 20 loci used accounting for only 65% of the variance.   Do you have access to the real #? 

Bob - In the Genetica paper out of UC Davis, they use 15 loci and the variances they define are balanced on average.

I am not sure what you are asking. I can not find a statement about % variance associated with a collection of SSR (microstaellite) loci. If you are asking whether a few alleles within any loci provide the large majority of the variance for that loci, that is true for some loci & not others. Figure 3 in this paper helps demonstrate that:
http://versita.metapress.com/content/572m54054284257l/fulltext.pdf

If you can point me to where you saw statements about % variance, I am happy to give my take on it.

Way over my head. Just trying to read, gives me a head ache. When they come out with fig genetics for dummies, I'll buy it. I am glad that there are members that understand this because it is important info.

luke

A little over my head too, but interesting, never the less. What, if anything, does this discussion  have to do with what some call "dark matter" in DNA strings that I have read act as or influence triggers of gene expression?

I guess it is a little like peeling the onion. We do DNA testing and check on almost infinitesimal amount it, which we have determined is enough to establish sameness of two different varieties. That was the first several layers: discovering "genetics", then chromosomes, then genes, then DNA. Now we are at the point of recognizing that the stuff we thought was important and unique (certain alleles), probably has other "stuff" that affects how it behaves or is "expressed" quite possibly based on environmental factors. The "dark matter" or "trash" as it has also been called, may be there for a reason - so we'll have to peel some more layers before we truly understand - or maybe that will reveal some more layers that we don't even know exist.

The problem with where we are at now is that our understanding of the sameness of the DNA doesn't match our empirical observations of what actually happens - that is, that variants with the "same" DNA don't look of act the same. That means that there are more layers to discover = more complexity to sort out.

In a different realm, but with the same issues and questions you have this: take a fertilized egg, and it differentiates into skin, bone, muscle, liver, heart, etc. They all have the same DNA, but something directs the following generations of cells to become different things (differentiate). How does that happen when the DNA is the same?

Jon - That is well said. As you said above, while the level of knowledge & understanding is constantly growing, it is a work in progress.

The terrific step forward here is that the tools now exist to compare any fig to this data set. So, if you don't know what your black fig is, you can take a couple of leaves and have someone type the DNA and compare the patterns from your fig directly to the NCGR collection. Yours may be the same as one there or a novel fig. The DNA will tell you quickly & fairly cheaply. If the DNA test says your fig is different from anything in the NCGR collection, it is a novel fig. If it is identical to a fig in the NCGR collection, then some of the issues we are discussing above come into play. Is it completely identical throughout all of its DNA? Are there epigenetic effects that make a difference in the two figs? A further point is that, as these test methods become standardized, figs from all the collections around the world can be compared. Once you know which Turkish or Tunisian or Spanish or Greek or French, etc figs are unique and which are the same as those in the NCGR, then people in this group are really in a position to gather some serious fig collections!!!

Good luck with your trees!

Jon,

As I tell my lab students "yeast and humans have 98% the same DNA". After all that is why we use yeast cells to experiment on to understand the human system. 

Yes and the yeast cells use us.

I am more confused now. What I'm reading seems to be worrying about naming figs. Even if you had a true name for a fig, that same fig has 10 or more other names that it has gone by for however long. Knowing the name will never change the flavor of the fig, nor will having a novel fig if the taste is not there. Please forgive me here, just trying to understand a little bit. When I think of DNA, I'm thinking getting down to the nitty gritty. Finding the good and the bad, and learn how to improve. Posible cure for FMV or
 one day or help with hybreds. Like I said earlier, it is over my head.

luke