These bags are already in the wild and underpins recent preprints showing strong perceptual validity:

pmc.ncbi.nlm.nih.gov/articles/PMC...
pmc.ncbi.nlm.nih.gov/articles/PMC...

Bottom line: jars leak information, bags keep it airtight. #bagboi #bagsrule

2/3 We’ve built and tested an odor sampling bag system with:
- Known concentrations
- Closed headspace
- Air as the carrier (no solvent confounds)

In head-to-head tests, bags produced more reliable results than jars and preserved odorant-specific differences.

We are only studying perception in this study with modeling informed by past molecular studies that have shown wide-spread inhibition for odor mixtures. One version of that inhibition is competitive binding as you refer to. See this paper: pubmed.ncbi.nlm.nih.gov/31000595/

This isn’t just theory!
These models can replace slow, expensive “omission tests” in flavor chemistry to find the handful of molecules that actually matter in a food’s aroma.
For instance, what molecules really make coffee smell like coffee. ☕

4/5
Using this information, we can predict perceived odor intensity from molecular structure + concentration with high accuracy for both:
• Single molecules (RMSE = 7.1 on a 100 point scale)
• Complex mixtures (RMSE ~ 7.5)

3/5
Using a very cool bag system to tightly control the concentration, we measured:
• 62 molecules, each at ≥7 concentrations
• 260 mixtures (2–10 components)
• >7,000 intensity ratings from trained human panels

2/5
It’s a trick question. At a lower concentration, acetophenone (left molecule, red, smells like urinal cake) is stronger than 2-heptanone (right molecule, green, smells like blue cheese). So concentration ≠ intensity. That’s been a major roadblock for neuroscience and fragrance / food industries

Hi, I'd like to be added to the feed to post a study:
scholar.google.com/citations?us...

work place: monell.org/robert-pelle...