how many organic compounds ?

Off-topic conversations and chit-chat.

26 posts Page 2 of 2
good points. I'd worry about conventional multivariate approaches applied to your problem anyway, because even if individuals don't always have a unique compound, your Dog-LOD is so very low that there may be relevant minor components that will be rejected as too variable, for example on the basis of too great an RSD in a mixed QC sample. The typical methods might be over-cautious; for (trivial) example, 8 ludicrously dilute chemicals detected by the dog as "present" or "absent" would already permit 255 individual smells. If an analyst doesn't bother looking at them because the RSD was worse than 20%, when the dog didn't really care at the 100% level, then he'll fail to understand what the dog is detecting...
The bottom line is that if someone wants to understand how a dog recognises other dogs, they have to look at the same data as the dog. That means that both the analytical equipment and the number-crunching methods have to work to the dilution the dog works. There I'd agree with you, even if you ask for that expensive equipment...
Good luck!
GOM wrote:
Hi CPG,

I meant identical. Like Peter I had an hypothesis which went against current thinking. Ultimately the only test of an hypothesis is by experiment. Unlike Peter I had the luxury of being able to put a human on the end of a GC column to sniff ( other workers in the insect field can attach an antenna to the outlet.) It took 3 years of hard work to confirm our hypothesis. I had to pull out every trick that I had ever learnt in of 30 years of specialising in GC and invent some new ones but we did it - it really did feel good when we proved it! :D

Cheers

Ralph


Hi Ralph - now I'm intrigued ! What was it that you proved ?

Peter
Peter Apps
lmh wrote:
good points. I'd worry about conventional multivariate approaches applied to your problem anyway, because even if individuals don't always have a unique compound, your Dog-LOD is so very low that there may be relevant minor components that will be rejected as too variable, for example on the basis of too great an RSD in a mixed QC sample. The typical methods might be over-cautious; for (trivial) example, 8 ludicrously dilute chemicals detected by the dog as "present" or "absent" would already permit 255 individual smells. If an analyst doesn't bother looking at them because the RSD was worse than 20%, when the dog didn't really care at the 100% level, then he'll fail to understand what the dog is detecting...
The bottom line is that if someone wants to understand how a dog recognises other dogs, they have to look at the same data as the dog. That means that both the analytical equipment and the number-crunching methods have to work to the dilution the dog works. There I'd agree with you, even if you ask for that expensive equipment...
Good luck!


Hi lmh

All of this is very true. The limitations of straightforward GC-MS analyses for mammal semiochemistry deserves a paper of its own, and will probably get one if I can find the time !

Peter
Peter Apps
... been thinking about it since last night. It's outside my field, so regard this as a question from an interested amateur: do you also have to do the synthetic bit? I.e. having characterised the chemicals that make up the odour of individual "A", do you then have to make up mixtures containing those chemicals and check that individual "B" still identifies your mix as "A"? The thought is that just because a chemical is present, that's no proof that it's part of the recognition process. It's a bit like geneticists' "promoter bashing" where they can't just hypothesise that conserved regions of promoter between a set of genes are probably involved in gene regulation, but must mutate/delete bits and see if the gene is still expressed. Must be horrendously complicated to do on complex smells? If I were atacking an LC problem like that, I'd be inclined to look at it as a bioassay, and reconstitute material from a complete run, minus a peak, to see if that peak is necessary for recognition...
Hi lmh

Considering that this is outside your field you talk more sense than some of the people whose field it is !!

Getting the target biological response to a synthetic mixture is the desired end point, and has been done in quite a few studies. But what also tends to happen is that an experiment is run that shows, say, that mice can discriminate old mice from young mice by smell. Then the odours of young and old mice are analysed, and consistent differences are found. The chemicals that account for this difference are mixed together, and an experiment is run to show that mice can discriminate between them. Then the thesis gets submitted, and the paper gets written. The final step, fooling a mouse into thinking that an old mouse is a young one and vice versa does not get done very often.

The whole area of mammal semiochemistry is pretty much a paradigm case of needing teams of top level specialists in biology, chemistry and data analysis, all working together and understanding one another. Just designing realistic bioassays is fraught with problems, and has led down a number of blind alleys. The collaboration and bioassay problems have been aired in the literature more than once.

Deleting peaks one at a time is logical (and I think that it has been used for flavours), but practically horribly difficult because of the complexity, and you would have to have a good straightforward bioassay to test the reconstituted mixtures. It would certainly be worth a shot with a rodent model to see whether it could work or not - its huge advantage would be that it would preserve all the multicompound signals until one of their members was cut out. It would also definitely exclude sections of the chromatogram whose deletion made no difference, and if those sections were made quite wide you could do a preliminary screening in only a few runs. You might be on to something !!

Peter
Peter Apps
Peter Apps wrote:
Another thing to keep in mind is that the LOD of a dog's nose gets down to 10 to the minus 18 molar


Hi, Peter. Since we are "around the water cooler," I suppose I ought to come out of lurk mode and mention that I took this quote and posted it on my facebook page, which is almost entirely about my hobbies/family, and nothing, other than this quote, from my professional life. One of the hobbies I started about a year ago is K9 Nosework with my two dogs. So I posted this quote. Then a friend of mine with no chemistry background at all asked what "molar" was, and proceeded to repost a modified version of the quote on her FB page. She has a lot of followers, so this quote is now rapidly spreading through the nosework community.

So, in case this quote ever gets tracked back to me, is there a reference in the literature for it? :)
All standard disclaimers apply. My posts are my opinions only and do not necessarily reflect the policies of my employer.
Ouch, what a difficult subject! Fascinating though... I shall follow it with interest. This thread has opened a window for me, and let a dog's nose in.
MaryCarson wrote:
Peter Apps wrote:
Another thing to keep in mind is that the LOD of a dog's nose gets down to 10 to the minus 18 molar


Hi, Peter. Since we are "around the water cooler," I suppose I ought to come out of lurk mode and mention that I took this quote and posted it on my facebook page, which is almost entirely about my hobbies/family, and nothing, other than this quote, from my professional life. One of the hobbies I started about a year ago is K9 Nosework with my two dogs. So I posted this quote. Then a friend of mine with no chemistry background at all asked what "molar" was, and proceeded to repost a modified version of the quote on her FB page. She has a lot of followers, so this quote is now rapidly spreading through the nosework community.


So, in case this quote ever gets tracked back to me, is there a reference in the literature for it? :)


The actual figure given is 10 e-17.67 M, as the threshold for alpha ionone in:

Passe, D.H and Walker, J.C. 1985. Odor psychophysics in vertebrates. Neuroscience and Biobehavioural Reviews 9: 431 - 467.

This is marginally better than rats' thresholds for most compounds. There is no reason to suppose that alpha ionone has any special significance for dogs, and it would be perfectly reasonable to assume that there are some compounds for which their thresholds are lower. On the other hand this is a single figure that most studies do not replicate - it is far more common for dog thresholds to be in the 10e-12 to 10e-14 range and above, but again these tests are run with abitrary compounds like amyl acetate.

Surprisingly enough, given the use that is made of sniffer dogs, there is surprisingly little hard science on what they can do and how they do it. In the tracker and search dog literature all the references ultimately trace back to non-peer-reviewed sources. Training and testing methods have a huge effect, of course, and there is the ever-present handler effect:

Lisa Lit • Julie B. Schweitzer • Anita M. Oberbauer 2011. Handler beliefs affect scent detection dog outcomes. Animal Cognition DOI 10.1007/s10071-010-0373-2.

What dogs do you have and what kind of scent work are they doing ?

Peter
Peter Apps
Hi, Peter.

Thank you for your response, and especially for the references. The 2011 one is very interesting, but not at all unexpected.

I have Belgian Tervuren, the long-haired version of Malinois, the breed with the largest representation in the paper you listed. We do nosework for fun, and are novices at it. The scent we use right now is birch. The training is really two-fold--teach the dog to recognize the scent, but also teach the handler to recognize when the dog has found the scent. I have been told that many top scent-detection dogs actually have minimal or no formal obedience training, because that kind of training encourages the dog to be dependent on the handler, and for scent-work one wants the dog to be operating independently of the handler.

FWIW, the authors of this paper used a difficult test. "Clearing" a location--saying that no target scent is there--is one of the more difficult tasks in nosework. It requires the handler to have faith in the dog, and most humans are just not that trusting. :wink: Sneaky research scientists lying to the gullible handlers! I bet next time the handlers will know to trust their dogs instead of the scientists!
All standard disclaimers apply. My posts are my opinions only and do not necessarily reflect the policies of my employer.
Hi Mary

I used to do ring obedience (with a staffordshire terrier !) and there were an astonishing number of dogs that would freeze on the correct scent, waiting for the handler to confirm with a recall. In training it was incredibly difficult to get people to allow the dog to get on with what it is good at, without constant "encouragement" and extra commands.

Have you seen the recent paper on the collie that learned over 1000 names for different objects ? - I thought that I had a copy but I cannot find it.

Peter
Peter Apps
I have revisited this because I am doing yet another review paper. It seems that the potential number of compounds is seriously mind numbing - http://ns1.peter-ertl.com/reprints/Ertl ... 4-2003.pdf and http://pipeline.corante.com/archives/20 ... pounds.php

Peter
Peter Apps
26 posts Page 2 of 2

Who is online

In total there is 1 user online :: 0 registered, 0 hidden and 1 guest (based on users active over the past 5 minutes)
Most users ever online was 1117 on Mon Jan 31, 2022 2:50 pm

Users browsing this forum: No registered users and 1 guest

Latest Blog Posts from Separation Science

Separation Science offers free learning from the experts covering methods, applications, webinars, eSeminars, videos, tutorials for users of liquid chromatography, gas chromatography, mass spectrometry, sample preparation and related analytical techniques.

Subscribe to our eNewsletter with daily, weekly or monthly updates: Food & Beverage, Environmental, (Bio)Pharmaceutical, Bioclinical, Liquid Chromatography, Gas Chromatography and Mass Spectrometry.

Liquid Chromatography

Gas Chromatography

Mass Spectrometry