Phylos Galaxy - Landrace discussion

mexcurandero420 said:

Only what i see is Purple Hz a pheno of Oldtimer Hz.Never heard of a purple Colombian Gold, but indeed of a purple Corinto Colombian.

Click to expand...

JGL's 72 Colombian Gold

Bio by Snowhigh:

http://genetixgenius.com/landrace-cannabis-strains/south-american-strains/colombian-heirloom-and-landrace-strains/colombian-gold/

Yes purple types do exist!

Note the 1980 Crosby Gold Colombian (on phylos) may be different to JGL '72 but is the parent of OTH without a doubt.

Also Crosby 1980 Gold Colombian can trace at least some parentage to 1970's commercial thai stick hence the Cambodian connection.

ngakpa said:

Phylos... why anyone places weight on it I don't know

Click to expand...

How anyone can't understand the basis of DNA paternity/maternity testing and it's application to phylos is beyond me.

Snowhigh is good at marketing bs

ngakpa said:

Phylos... why anyone places weight on it I don't know

Click to expand...

because they have an understanding of gentic testing and mapping

Nexus7 said:

How anyone can't understand the basis of DNA paternity/maternity testing and it's application to phylos is beyond me.

Click to expand...

you seem a touch... invested...

anyway, it's clear Phylos itself has revealed its own limitations just in its results...

never mind the hermeneutic issues involved when laymen with zero understanding of plant genetics - never mind cannabis genetics - start treating it as plain gospel... which is what's happening

ngakpa said:

Phylos has limitations on multiple levels, such as the doubtful authenticity of the material used, obviously dodgy results, and the fact that it's aimed at laymen despite the fact they/we are the last ones in a position to understand the inherent limitations of the results or how to interpret them

case in point is how people are now proclaiming that Old Timer's Haze is definitely a pure Colombian, or that Nepali and Thai cannabis are definitely closely related

if you showed this to a qualified plant geneticist or botanist do you think they'd be nodding along in agreement?

or look at some of the obviously absurd results:

take the Yarkhun and Laspur seeds that Mriko collected and gave to Sam. I was in Pakistan with him and met him on his way back from Chitral. I even know where the farmer he got some of the seeds from lives. So I can have good confidence that they are what Mriko says they are.

Yarkhun and Laspur are connected valleys. The Yarkhun is from Tehsil Mastuj. The Laspur were from mere kms distant. Look on Google Maps. Far as local farmers are concerned, they're the same plant.

Yet if you look at Phylos, apparently Senegali cannabis is more closely related to one (Yarkhun iirc) than Yarkhun and Laspur are to each other. There was a user on here defending this result on the grounds that Phylos must be right lol

So yeah, I'm not taking Phylos too seriously, and still less how it's being interpreted by non-specialists

Click to expand...

Hi
The Yarkuns, the Laspur and the Senegalese are all grouped fairly closely on Phylos Galaxy. When the Laspur and Yarkhun samples were seeds, it just could be that the individual seeds were somewhat distant to one another. Landrace populations aren't stable, am sure you are well aware of this.

Atm, the "Closest Genetic Relatives"-tables aren't working on Phylos for many, many samples, and it has been that way since they started updating the platform. None of the samples mentioned show any relation to other plants; not even the two Yarkhuns show each other in the relatives-table, but you can see them fairly close to each other in the "Gallaxy". And like i mentioned, the Laspur and Senegalese samples are also fairly close to the Yarkhuns in the Galaxy-view.

I don't mean to be an arse, but are you sure you understand fully well how Phylos works, cause when i look at the samples mentioned, i think you're making too big of a deal out of this. = the Senegalese sample originally comes from Pakistan, maybe even from the Yarkhun area.

ngakpa said:

you seem a touch... invested...

anyway, it's clear Phylos itself has revealed it's own limitations just in its results...

never mind the hermeneutic issues involved when laymen with zero understanding of plant genetics - never mind cannabis genetics - start treating it as plain gospel... which is what's happening

Click to expand...

No I'm not invested at all. I'm just interested in tools like phylos and kannapedia as an end level consumer.

There are too many tall and conflicting stories in regards to providence of certain "strains".

DNA testing can add some certainty about whose work gave rise to whose as well as information regarding homozygosity of a strain as well as other traits.

GoatCheese said:

Hi
The Yarkuns, the Laspur and the Senegalese are all grouped fairly closely on Phylos Galaxy. When the Laspur and Yarkhun samples were seeds, it just could be that the individual seeds were somewhat distant to one another. Landrace populations aren't stable, am sure you are well aware of this.

Atm, the "Closest Genetic Relatives"-tables aren't working on Phylos for many, many samples, and it has been that way since they started updating the platform. None of the samples mentioned show any relation to other plants; not even the two Yarkhuns show each other in the relatives-table, but you can see them fairly close to each other in the "Gallaxy". And like i mentioned, the Laspur and Senegalese samples are also fairly close to the Yarkhuns in the Galaxy-view.

I don't mean to be an arse, but are you sure you understand fully well how Phylos works, cause when i look at the samples mentioned, i think you're making too big of a deal out of this. = the Senegalese sample originally comes from Pakistan, maybe even from the Yarkhun area.

Click to expand...

I'm not sure I've understood what you mean by "the Senegalese sample originally comes from Pakistan, maybe even from the Yarkhun area" because the only reading I can make of that is daft as

last I looked at the Galaxy, it clearly indicated that iirc the Yarkhun is more closely related to an accession from Senegal, West Africa than it is to the Laspur accession, depite the fact that Laspur village is literally a few kms drive away from Mastuj, which is where the Yarkhun accession is from... (Laspur Valley literally joins Yarkhun Valley at Mastuj village)

Please telll me you aren't saying that you think the Senegali landraces come from Yarkhun... right?

Nexus7 said:

No I'm not invested at all. I'm just interested in tools like phylos and kannapedia as an end level consumer.

Click to expand...

but of course...

GoatCheese said:

Atm, the "Closest Genetic Relatives"-tables aren't working on Phylos for many, many samples, and it has been that way since they started updating the platform. None of the samples mentioned show any relation to other plants; not even the two Yarkhuns show each other in the relatives-table

Click to expand...

Right, so, based on what you're saying, Phylos is in a right fecking mess...

ngakpa said:

I'm not sure I've understood what you mean by "the Senegalese sample originally comes from Pakistan, maybe even from the Yarkhun area" because the only reading I can make of that is daft as

last I looked at the Galaxy, it clearly indicated that iirc the Yarkhun is more closely related to an accession from Senegal, West Africa than it is to the Laspur accession, depite the fact that Laspur village is literally a few kms drive away from Mastuj, which is where the Yarkhun accession is from... (Laspur Valley literally joins Yarkhun Valley at Mastuj village)

Please telll me you aren't saying that you think the Senegali landraces come from Yarkhun... right?

Click to expand...

Yea, i claimed that as a fact, didn't i. Grow up.

No genetics from India/Pakistan travelled to Africa during the British India times, when lots of Indians worked in Africa and Trade Company ships sailed all over? How daft does that claim sound when you think about it.

ngakpa said:

Right, so, based on what you're saying, Phylos is in a right fecking mess...

Click to expand...

You didn't know Phylos platform hasn't been working very well because of the update before i told you about it? Yea, let's only listen to you, cause you've visited Pakistan.

No genetics from India/Pakistan travelled to Africa during the British India times, when lots of Indians worked in Africa and Trade Company ships sailed all over? How daft does that claim sound when you think about it.

Click to expand...

They speak French in Senegal and for a long time the Arabs ruled overthere before the French came, so the connection between Senegal and Pakistan is a bit strange to me.

I wonder who else is working on similar projects and how the results compare? To me the weakest link in the phylos is that people need to both know and be honest about there sample (might be tricky for some). I would think Arjan has a similar database although I cant imagine him making the results public any more than he does with his seeds.

mexcurandero420 said:

Only what i see is Purple Hz a pheno of Oldtimer Hz.Never heard of a purple Colombian Gold, but indeed of a purple Corinto Colombian.

Btw don't know how accurate they are, but the direct link of Colombian Gold to Cambodian from Mel Frank makes me think if Mel Frank made a mistake or they are.

Click to expand...

What people were smoking in the States and elsewhere were named by the appearance of the buds, the purple ones would be called colombian red and the green ones gold. The local names for the actual plants would have been punto rojo or mangobiche for example.


The one Colombian sample from 1920's, the colombian reds and mangobiche all sit close to each other in the Galaxy while Oldtimer haze and the Colombian Gold samples form a branch of their own further from the others.



At face value it looks like Purple Colombian is just a phenotype. Mangobiche is the name given to the old Colombian line that looks derived from the 1920's sample. Colombian Red, punto rojo, moño rojo, red corrento (describing phenotypes) all sit close to each other and it's not a coincidence.



I'm not saying Phylos is gospel.. I'm saying it's the best thing we have and it's giving better matches all the time. If you know how to interprete the results it can give some valuable insight.



I'll get back to this later.

Ngakpa, I appreciate the dose of healthy scepticism, Phylos has it's problems.



The most obvious is their using modern cultivars as reference groups. When a 1920's Colombian (link) shows "berry" ancestry you can be sure it's not referring to any modern hybrid. It can also give the impression that a 1979 Oaxacan is derived from skunk when opposite is more likely to be true.



We do not know which snippets of dna they chose to represent which cultivar. How this works is you choose a number of genetic markers that you deem specific for the population or individual in question and then compare the results with others. The skunk markers might or might not account for both Mexican and Afghani specific dna or a certain combination of both. What the berry component stands for is beyond me.. It is too widespread in the Himalayas and Central Asia to be indicative of contamination.



Let's take Skunk#1 as an example. This plant is from the Skunkman himself, we know what it's ancestry proportions should look like. 100% skunk, right?


Instead we get this (link)



It still looks like a 85% match and the rest can be interpreted as being the margin of error we have to deal with at this point. When we're looking at landraces that can be interpreted as little more than "noise" but in the question of Skunk we can assume that it does actually contain ancestry similar to hemp, og kush, cbd.. those markers just arent present in their skunk sample and their method interprets the results like this.



So there are things Phylos can and can't do very well. What I think it does do pretty well is show the contamination of landraces by modern hybrids because the Skunk and Ogkush have a very specific combination of markers that clusters them away from the larger landrace branch in the three dimensional Phylos galaxy. It also works very well in showing the proportions of the landrace (NLD) and hemp (very widespread) specific ancestry.



You shouldn't be seeing much if any of Skunk ancestry outside Central America, the OG component might also be a sign of contamination (or additionally geneflow from Central Asia) but berry-like ancestry seems to be present in many greater Himalayan landraces as well as Africa. This is probably an artefact of the markers used.. And most of the time ancestry proportions around or below 10% should especially be taken with a grain of salt as this fits in the margin of error we know exists, and is likely to be just "noise".



Another thing I think Phylos does a decent job at is showing immediate genetic relatives. Yes, there have been obvious mistakes but they keep piecing things togerther and the results you got 12 months ago will probably look different now.


Phylos is able to distinguish three levels of immediate relation. 1. A clonal relationship, 2. siblings of the same strain and 3. close relatives. If we look at the inbred Skunk#1 again we can see that both of the Skunk#1 samples fall in the second category. Oldtimer haze and Colombian are shown as bit further to each other but still close relatives, possibly due to not being IBLs.



I do not know why Phylos sometimes fails to connect closely related samples but when it does find the connection, I tend to trust it.



I looked at those Pakistani samples a minute ago and they now cluster close to each other in the galaxy and Senegal is further as it should, one just needs to rotate the view to see where they sit in 3D. Have you checked if it looks different to what it used to Ngakpa? Looking at the Senegalese (link) it's an obvious hybrid which might explain why it looks that way and does not cluster close to the other African strains. It might be admixed with a Pakistani strain or just has similar ancestry proportions (of landrace, hemp, skunk and berry) that makes it look genetically similar.



So, there are things Phylos at this point can do, and what it can't do. There are limitations that one has to take into consideration but if you spend enough time looking through the data there are certain patterns that I personally find very useful and during the couple of years the results have been getting more and more consistant.

mexcurandero420 said:

They speak French in Senegal and for a long time the Arabs ruled overthere before the French came, so the connection between Senegal and Pakistan is a bit strange to me.

Click to expand...

Well my point wasn't to claim it was the British who brought cannabis to Africa, thou i mentioned them, it was to point that Trade Companies (Brits, Dutch etc.) sailed around Africa and India.

Google East Indian Company Dakar, or something similar, and you'll see that many trade companies/European powers sailed thru there.
If Arabs or the French brought cannabis to Senegal it's fine by me. The French and the Brits fought wars in India/Afghanistan during the "Great Game", so there is a French connection to Africa and "India" if you need to find one.


To me it's quite amazing that someone thinks it's silly that "indian" genetics could have found their way to Africa, and that if Phylos-data shows it to be so then it has to be a mistake by them or dodgy information.

Phylos shows many connections of "Indian" genetics to cannabis in different African countries. Also alot of SE Asian genetics are in Africa according to Phylos.

GoatCheese said:

Well my point wasn't to claim it was the British who brought cannabis to Africa, thou i mentioned them, it was to point that Trade Companies (Brits, Dutch etc.) sailed around Africa and India.

Google East Indian Company Dakar, or something similar, and you'll see that many trade companies/European powers sailed thru there.
If Arabs or the French brought cannabis to Senegal it's fine by me. The French and the Brits fought wars in India/Afghanistan during the "Great Game", so there is a French connection to Africa and "India" if you need to find one.


To me it's quite amazing that someone thinks it's silly that "indian" genetics could have found their way to Africa, and that if Phylos-data shows it to be so then it has to be a mistake by them or dodgy information.

Phylos shows many connections of "Indian" genetics to cannabis in different African countries. Also alot of SE Asian genetics are in Africa according to Phylos.

Click to expand...

The Arabs were in the slave trade for 14 centuries before the Europeans came to Africa and there are Arab writings about the psychoactivity of Cannabis dated 16th century.

The Dutch grew hemp for the fibre around the Amsterdam area, but their trade was in spices from Asia.In the early 20th century there was Cannabis cultivation in Indonesia (Java) for the delivery to the pharmacies in the Netherlands.

Thule said:

Ngakpa, I appreciate the dose of healthy scepticism, Phylos has it's problems.

The most obvious is their using modern cultivars as reference groups. When a 1920's Colombian (link) shows "berry" ancestry you can obviously discard that as noise. It can also give the impression that a 1979 Oaxacan is derived from skunk when opposite is more likely to be true.

Click to expand...

I find it abit strange how many people seem to be stuck on the "Skunk/Landrace/berry/what ever" chart on Phylos, but that's just me. To me it's quite pointless as you rightly say, but cause it does seem so pointless then why even think about it, you know??
Me, i look thru the "genetic relatives tables" (when those were working) or the so called Galaxy, and i pretty much pay no attention to the "Skunk/Landrace/berry/what ever" chart.

Thule said:

I do not know why Phylos sometimes fails to connect closely related samples but when it does find the connection, I tend to trust it.

Click to expand...

The connecting-thing started going dodgy when they updated the platform over a year ago. I was surfing the site quite abit before it went shitty, and the "connecting" worked fine back then. Many samples that showed connections in the "Genetic Relatives"-table then aren't showing them anymore.

Here's one example:
https://phylos.bio/sims/sample/genotype/eg5r66v8

This is screen capture of X18 sample before the Phylos update:




Here's how the page looks today:

GoatCheese said:

I find it abit strange how many people seem to be stuck on the "Skunk/Landrace/berry/what ever" chart on Phylos, but that's just me. To me it's quite pointless as you rightly say, but cause it does seem so pointless then why even think about it, you know??
Me, i look thru the "genetic relatives tables" (when those were working) or the so called Galaxy, and i pretty much pay no attention to the "Skunk/Landrace/berry/what ever" chart.



The connecting-thing started going dodgy when they updated the platform over a year ago. I was surfing the site quite abit before it went shitty, and the "connecting" worked fine back then. Many samples that showed connections in the "Genetic Relatives"-table then aren't showing them anymore.

Here's one example:
https://phylos.bio/sims/sample/genotype/eg5r66v8

This is screen capture of X18 sample before the Phylos update:

View Image


Here's how the page looks today:


View Image

Click to expand...

It is all about missing links and honesty. If everyone had saved a sample of everything and it was always what it was supposed to be and we all knew where it came from this system would be amazing. And I thinks as time goes onit will improve but there is alot of room for misinformation to go unnoticed.

GoatCheese said:

I find it abit strange how many people seem to be stuck on the "Skunk/Landrace/berry/what ever" chart on Phylos, but that's just me. To me it's quite pointless as you rightly say, but cause it does seem so pointless then why even think about it, you know??
View Image

Click to expand...

Oh I didn't mean it's pointless, not at all. It tells very little about ancient cultivars and modern hybrids but it's my favorite thing to look at when trying to figure out the composition of landraces. The presence of the skunk component in old world landraces should be very useful in identifying contamination in contemporary landraces. You just have to know how to work your way around Phylo's shortcomings.