Neuromorphic Odor Translator Helps Robots Express Their Feelings

Neural network trained to properly name organic molecules

Aug 2021, phys.org

https://techxplore.com/news/2021-07-neural-network-properly-molecules.html

Do you ever have a problem naming that smell? It's not just you. Science has this problem too, but maybe not for long. 

Smells are volatile organic compounds that have evaporated and entered your nose. Although they almost always occur in combination with others and not in isolation, us humans want to reduce smells to their individual components, and then name them. After all, in order to think about something, you have to know it's name. (Is that true?) 

The problem is that organic molecules are big, with lots of chemicals joined together in lots of ways, so coming up with a naming convention for all these permutations is hard. IUPAC, the International Union of Pure and Applied Chemistry, sets the convention for naming molecules. And boy is it complicated.

Take sugar, a common molecule known to us by its simple name "sucrose;" in IUPAC, it's called (2R,3R,4S,5S,6R)-2-[(2S,3S,4S, 5R)-3,4-dihydroxy-2,5-bis(hydroxymethyl)oxolan-2-yl]oxy-6-(hydroxymethyl)oxane-3,4,5-triol.

Since we do live in the computer age, folks want to automate this naming process for when they discover new molecules. But as you can imagine by looking at the IUPAC name for sucrose, the algorithm at the core of that naming convention is really hard to write. So they decided to use a neural network instead.*

*I'm casually calling a neural network "neuromorphic," but in the past few years, real neuromorphic computers have forced a distinction here that I'm ignoring here for the sake of a more clickable title. 

This new artificially intelligent chemical translator is not a magical structure-to-name translator that can just look at a chemical and give it a name; that's still out of reach. It does, however, translate between IUPAC and another naming convention called SMILES.

Trained on PubChem's 100 million molecules, this translator ultimately shows how the utility of the new approach of using neural networks to help us write algorithms from the bottom up instead of the top down, which really is a revolution in computing. 

And if you think it would be cool to have robots that can smell, or to ensure that future humans maintain their sense of smell as they evolve into hyperdimensional algorithms, then making odors machine-readable is how you do that.

via Skolkovo Institute of Science and Technology, Lomonosov Moscow State University and start-up Syntelly: Lev Krasnov et al, Transformer-based artificial neural networks for the conversion between chemical notations, Scientific Reports (2021). DOI: 10.1038/s41598-021-94082-y

Dark Taxa AKA Creating Taxonomies From Scratch

New norms needed to name never-seen fungi

May 2021, phys.org

https://phys.org/news/2021-05-norms-never-seen-fungi.html

There's 150,000 species of fungi known, yet a projected 2.2 to 3.8 million still waiting to be discovered (these are called dark taxa). But because of advances in DNA sequencing and microscopy, we're learning so fast that we need a new way to organize it all. 

This comes up in the context of biosecurity, where it can only work if "organisms detected can be reliably identified and have accurate names." For fungi, that's not really possible, because believe it or not, we don't have a good catalog of fungi. 

-via: Robert Lücking et al. Fungal taxonomy and sequence-based nomenclature, Nature Microbiology (2021). DOI: 10.1038/s41564-021-00888-x

We also don't have a good way to organize the words we use to describe everyday smells, and we don't have something like a "smell taxonomy." There are plenty of sub-domains that organize their relevant smells, found in subjects like coffee, wine, perfume, and culinary arts. They always seem to take the form of a wheel (not the most complex form). You can get a good start with everyday smells at the South Coast Air Quality Management District, who created a "Characterization of Odor Nuisance" odor wheel, with the help of environmental scientist Jane Curren at UCLA circa 2016. It was based on a bunch of phone calls made to the District where people were complaining about odors in their neighborhood. She took all the words they used and organized them. 

You could also look into Ann-Sophie Barwich who is a cognitive scientist who did her dissertion on olfactory categorization, and then wrote a book called Smellosophy. Probably one of the most interesting academics you will ever hear of. I mean, her master's thesis was about the relevance of  Leibniz causality on biological classification.

Image credit: Penicillin, Kew Royal Botanical Gardens for BBC

Notes:

State of the World's Fungi, by the Kew Royal Botanical Gardens (2018), is the first ever State of the World's Fungi report revealing how important fungi are to all life on Earth. [pdf]

[State of the World's Fungi]

International Commission on the Taxonomy of Fungi (ICTF)

MycoBank is the on-line repository and nomenclatural registry provided in collaboration between the International Mycological Association and the Westerdijk Fungal Biodiversity Institute. It provides a free service to the mycological and scientific society by databasing mycological nomenclatural novelties (new names and combinations) and associated data, such as descriptions, illustrations and DNA barcodes. Nomenclatural novelties are each allocated a unique MycoBank number to be cited in the publication where the nomenclatural novelty is introduced, to conform with the requirements of the International Code of Nomenclature for algae, fungi and plants.

Identification and quantification of nuisance odors at a trash transfer station. Jane Curren, et al.  PubMed, Waste Manag. 2016 Dec;58:52-61. doi: 10.1016/j.wasman.2016.09.021. Epub 2016 Sep 28.

https://pubmed.ncbi.nlm.nih.gov/27692531/

Post Script:

I'm looking at a popular science article about fungi. The first two "interesting" points, when looked at together, remind me of why I always have the feeling like fungi are from outerspace:

• Fungi are in a kingdom of their own but are closer to animals than plants
• They have chemicals in their cell walls shared with lobsters and crabs (you do know we're all becoming crabs, right?)

Semantic Bingo

In olfactory research, there's a test called a pairwise similarity test that's used to measure smells, allowing researchers to construction of a map of odor perception.

It's hard to make sense out of smells; it doesn't work like the rest of our sensory system. For a bunch of reasons it's proven quite difficult to produce a model which predicts how a molecule will be perceived.

It's not broken beyond repair, but it is frustrating because we can never seem to get an airtight model that works for all smells and for all people. With hundreds of different receptors, varying over thousands of alleles, scientists often look somewhere else for the organizing principles -- they look for patterns in the words themselves.

In a study from 2015, distributional semantics is used to create an odor map. They say it's the first attempt to do so. This technique rests on the theory that words occuring in similar contexts are in fact similar. Some of you might remember this as "context clues;" if you come across a new word while you're reading, use the surrounding context to help you guess what the word means.

So instead of trying to make a map of molecular features and receptor actuation potentials, they make a map of the words themselves. They use large text datasets, i.e., really big books, one of which was the Sigma-Aldrich Flavors and Fragrances catalog, then score words based on their co-occurances in the text.

I started this post just so I could paste these lists of words, so let's get on with it. On a scale of 0-1, how likely is it that these words can be interchanged?

Similarity Test:

bakery-bread  0.96

grass-lawn       0.96

dog-terrier      0.90

bacon-meat    0.88

oak-wood        0.84

daisy-violet     0.76

daffodil-rose   0.74

Nearest Neighbor Test:

apple - pear, banana, melon, apricot, pineapple

bacon - smoky, roasted, coffee, mesquite, mossy

brandy - rum, whiskey, wine-like, grape, fleshy

cashew - hazlenut, peanut, almond, hawthorne, jam

chocolate - cocoa, sweet, coffee, licorice, roasted

lemon - geranium, grapefruit, tart, floral

cheese - grassy, butter, oily, creamy, coconut

caramel - nutty, roasted, maple, butterscotch, coffee

Notes:

Kiela, D., Bulat, L. & Clark, S. Grounding semantics in olfactory perception. Assoc. Comput. Linguist. 231–326 (2015).

https://www.aclweb.org/anthology/P15-2038.pdf

Distributional Semantics – represents the meanings of words as vectors in a “semantic space”, relying on the distributional hypothesis: the idea that words that occur in similar contexts tend to have similar meanings.

Local Odor Vocab

Curren's SCAQMD Urban Odor Dictionary

link to google fusion graph

You’re looking at a smell network based on the work of UCLA post-grad named Jane Curren. She took all the words used in odor complaints in southern California, and found the source of those complaints. These odor sources can be anything from a hidden garbage dump to a local restaurant. The size of the blue nodes doesn’t indicate smelliness; it indicates more descriptors. Restaurants have more descriptors because they have 1. More people near them and 2. A greater diversity of things that smell. (I’m guessing this.) What this chart does show is the most common odor complaints in the middle of the cluster – we smell burnt things a lot. Also rotten eggs/sulfur/natural gas.

Below, Curren went further and found the specific chemical source of these smells, and I listed them below. Next time you smell something funny, check this list. (And the next time you’re in New Jersey, take a tour of the New Jersey Turnpike mobile museum of olfactory delights, you’ll smell firsthand many of the odors on this list!)

Curren's Urban Odor Lexicon

               

Descriptor                           Odorant

solvent                                 2-butanone

petroleum                          2-methyl-1-propene

pungent                               2-pentanone

sweet                                   2-pentanone

lemon                                   acetaldehyde

alcohol                                  acetaldehyde

acrid                                      acrolein

pungent                               ammonia

sweet solvent                   benzene

sweet                                   dichloromethane

decayed cabbage             ethyl mercaptan

pungent                               formaldehyde

rotten egg                           hydrogen sulfide

woody                                  m-cresol

resinous                               m-cresol

medical                                                m-cresol

woody                                  o-cresol

resinous                               o-cresol

medical                                                o-cresol

woody                                  p-cresol

resinous                               p-cresol

medical                                                p-cresol

medical                                                phenol

sweet                                   phenol

irritating                               propanal

fruity                                     propanal

empyreumatic                  pyridinec

irritating                               sulfurdioxide

aromatic                              toluenec

fishy                                      trimethylamine

pungent                               trimethylamine

sweet                                   acetone

minty                                    acetone

sweet solvent                   benzene

woody                                  cresol

resinous                               cresol

medicinal                             cresol

sour                                       dimethyl disulfide

onion                                    dimethyl disulfide

decayed cabbage             dimethyl sulfide

decayed cabbage             ethyl mercaptan

garlic                                      ethyl mercaptan

aromatic                              ethylbenzene

pungent                               formaldehyde

rotten eggs                         hydrogen sulfide

skunk                                    i-propyl mercaptan

sour                                       methyl mercaptan

garlic                                      methyl mercaptan

decayed cabbage             methyl mercaptan

skunk                                    n-butyl mercaptan

skunk                                    n-propyl mercaptan

medicinal                             phenol

sweet                                   phenol

irritating                               propanal

sweet                                   p-xylene

sharp                                     thiophene

skunk                                    thiophene

rubber                                  toluene

moth balls                           toluene

aromatic                              1,2,4-trimethylbenzene

aromatic                              1,3,5-trimethylbenzene

vinegar                                 acetic acid

sour                                       acetic acid

chemical                              acetone

sweet                                   acetone

acrid                                      acrolien

pungent                               ammonia

sweet                                   a-pinene

pine                                       a-pinene

sweet solvent                   benzene

malty                                     butanal

burnt                                     butanal

rancid                                    butyric acid

sour                                       butyric acid

perspiration                       butyric acid

fruity                                     butyl acetate

dead body                          cadaverine

disagreeable sweet        carbon disulfide

almond                                 chlorobenzene

sweet                                   chloroform

ethereal                               chloroform

pungent                               crotonaledehyde

gasoline                               decane

ethereal                               dichloromethane

sour                                       dimethyl disulfide

onion                                    dimethyl disulfide

decaying vegetation       dimethyl sulfide

aromatic                              ethylbenzene

oily                                         heptanal

musty                                   heptanal

woody                                  heptanal

fatty                                      hexanal

green                                    hexanal

rotten egg                           hydrogen sulfide

sharp                                     isopropyl benzene

aromatic                              isopropyl benzene

sharp                                     cumene

aromatic                              cumene

lemon                                   limonene

decayed cabbage             methyl mercaptan

moth balls                           naphthalene

tar                                          naphthalene

sweet                                   m-xylene

gasoline                               octane

sweet                                   o-xylene

pungent                               pentanal

sharp                                     propanal

vinegar                                 propanoic acid

sour                                       propanoic acid

sweet                                   p-xylene

solvent                                 styrene

rubber                                  styrene

sweet                                   tetrachloroethene

rubber                                  toluene

moth balls                           toluene

sweet                                   tricholoroethane

fishy                                      trimethylamine

fecal                                      valeric acid

sour                                       valeric acid

sweet                                   vinyl acetate

Suffet and Rosenfeld's Urban Odor Lexicon

*much of Curren’s work came from Suffet and Rosenfeld, so I added their leftover terms here

Descriptor                           Odorant

Coffee                                  Furfruryl Thiopropionate

earthy                                   Geosmin

musty                                   2-Methyl isoborneol

moldy                                   2,4,6-Trichloroanisole

grassy                                   Cis-3-hexen-1-ol

dead animal                       Putresine

rotten vegetable              Dimethyl Trisulfide

fishy                                      Dimethyl Amine

fishy                                      Methyl Amine

plastic                                   Methyl Methacrylate

fecal                                      Indole

manure                                  Skatole

burnt                                     Guiaicol

Primary document: Curren, J. 2012. Characterization of Odor Nuisance. UCLA.

http://www.wcsawma.org/wp-content/uploads/2012/11/JMC-dissertation.pdf

Supporting document: Suffet, I.H., and P.E. Rosenfeld (2007). The Anatomy of Odour Wheels for Odors of Drinking Water, Wastewater, Compost and the Urban Environment, Water Science and Technology 55(5), 335-344.

Research note: The descriptor names come from the above source and supporting documents, whereas the odor causing compounds were matched against varying sources which will not be listed here; see instead the source document for those references