Project ID: 16736 | Inventors: Dr. Shayma El-Atawneh & Prof. Amiram Goldblum | Molecular Modeling and Drug Design Lab, Institute for Drug Research, The Hebrew University of Jerusalem
| Category | Computational Chemistry | Food Technology Machine Learning |
| Keywords | Sweetener discovery, taste prediction, low-calorie ingredients, bitter aftertaste, food science, receptor binding |
| Development Stage | TRL 2–3 — Computational proof-of-concept; 92 novel candidates identified and prioritized for experimental validation |
| Collaboration | Sponsored R&D | License Option | Co-Development Partnership |
| A proprietary discovery platform that identifies novel high-intensity sweetener candidates free of bitter aftertaste — delivering a curated library of 92 experimentally ready lead compounds with receptor-binding profiles that match or exceed the best FDA-approved sweeteners on the market today. |
Background & Market Opportunity
The global artificial sweeteners market is projected to exceed $29 billion by 2026 and $40+ billion by 20301, driven by demand for sugar reduction across food, beverages, nutraceuticals, and pharmaceuticals. Despite this growth, only six synthetic sweeteners have received full FDA approval in the past century5 — a pace entirely mismatched to the scale of the global health challenge.
The core obstacle is a dual problem that has defeated traditional screening for decades: finding a compound that is simultaneously intensely sweet and completely free of bitter aftertaste. Stevia, Saccharin, acesulfame-K, cyclamate, and even market-leading sucralose all carry measurable bitter off-notes. Mounting health concerns — including associations with gut microbiome disruption and elevated cardiometabolic risk4 — are further eroding trust in current options. The food and beverage industry spends over $32 billion annually on R&D2, yet a clean-tasting, safe next-generation sweetener remains elusive.
Competitive Landscape
Existing computational taste-prediction tools — including BitterMLP, BitterCNN, Premexotac, BitterPredict, and BitterSweetForest — address sweetness or bitterness in isolation and lack integration with receptor-structure data. Major flavor houses (Givaudan, IFF, Firmenich, Symrise) invest heavily in taste modulation, but no commercially validated, patentable novel sweetener candidates have yet emerged from AI-driven pipelines. This leaves a clear first-mover window for an industry partner that moves quickly.
Our Innovation
Researchers at The Hebrew University of Jerusalem have developed a proprietary ML-driven platform that screens vast chemical libraries to identify compounds predicted to be intensely sweet and simultaneously non-bitter, then validates top candidates against the human sweet taste receptor structure which is a novel protein structure published very recently (2025). The output is a curated, patent-pending library of 92 novel candidate molecules — ready to enter synthesis and sensory validation.
- Best-in-class predictive accuracy: The platform’s taste-classification models outperform other published methodologies — including BitterMLP-Fingerprint, BitterCNN, and Premexotac — on both accuracy and Matthews Correlation Coefficient (MCC)..
- First integrated sweet + anti-bitter pipeline: The platform simultaneously selects for sweetness while screening out bitter receptor activity — directly solving the aftertaste problem that disqualifies most known sweetener candidates.
- Receptor-validated hits: All 92 priority candidates were assessed for binding to the human sweet taste receptor (T1R2/T1R3), with docking scores that match or exceed those of advantame, the most potent FDA-approved sweetener (EC₅₀ = 0.9 µM). Top leads reach -8.8 kcal/mol vs. -7.8 kcal/mol for advantame.
- Synthesis-optimized leads: 43 of the 92 candidates have ≤1 chiral center — the molecular property most directly linked to manufacturing cost. Simpler stereochemistry typically reduces synthesis cost by 30–60% and shortens regulatory development timelines.
- Platform scalability: The[AG1] same technology generates novel bitter compound leads for the pharmaceutical taste-masking sector, doubling the addressable market with no additional platform development.
Commercial impact: A licensing partner gains a proprietary library of 92 novel candidate structures — each with receptor-interaction data in hand — compressing 3–5 years of traditional hit-identification campaigns into months. In the $40B+ sweetener market, patenting even one clean-taste high-intensity sweetener can command licensing fees in the tens of millions and multi-decade royalty streams across food, beverage, nutraceutical, and pharmaceutical applications.16
Results: Lead Candidates vs. FDA-Approved Controls
Using a proprietary computational platform, the team identified 92 novel candidates with predicted high-intensity sweetness and low bitter receptor activity. Top candidates were validated by assessing fit to the published human sweet taste receptor structure — the same binding site shared by all FDA-approved sweeteners. The table below compares receptor-binding profiles of the novel lead candidates against the seven FDA-approved synthetic sweeteners used as controls. A more negative docking score indicates stronger predicted binding and higher expected sweetness potency.
Table 1. Receptor Binding Comparison: Novel Lead Candidates vs. FDA-Approved Sweetener Controls
| Compound | Docking Score (kcal/mol) | EC₅₀ (µM) | H-Bonds | VdW Interactions | Buried SA (Ų) | Notes |
| Known Sweeteners — FDA-Approved Controls | ||||||
| Advantame | -7.8 | 0.9 | 5 | 426 | 1,046 | Most potent control |
| Neotame | -7.3 | 2.26 | 5 | 355 | 932 | Control |
| Sucralose | -7.2 | 80 | 5 | 237 | 744 | Reference standard |
| Aspartame | -6.5 | 750 | 6 | 224 | 781 | Control |
| Cyclamate | -5.6 | 256 | 3 | 149 | 463 | Control, FDA banned since 1970s |
| Acesulfame-K | -4.9 | 540 | 1 | 105 | 391 | Control |
| Saccharin | -4.8 | 190 | 0 | 138 | 442 | Weakest binder |
| Novel Candidates — Proprietary Platform Output ★ | ||||||
| Compound | Docking Score (kcal/mol) | EC₅₀ (µM) | H-Bonds | VdW Interactions | Buried SA (Ų) | Notes |
| Top 43 Priority Leads (≤1 chiral center) | -6.0 to -8.1 | Pending* | Up to 7 | <Advantame | < Advantame[Sש2] | ★ Exceed best control |
| Full Lead Library (92 candidates) | -6.0 to -7.3 | Pending* | 2–7 | > Sucralose | > Sucralose | ★ many exceed reference[Sש3] |
The novel lead candidates achieve docking scores up to -8.1 kcal/mol, surpassing advantame (-7.8 kcal/mol), the most potent known sweetener by receptor affinity. This suggests the leads may exhibit higher predicted sweetness potency than any currently approved synthetic sweetener. Experimental EC₅₀ values and sensory profiles will be determined in the next funded phase.
Partnership Opportunity
The novel sweetener candidates can be synthesized for the next phase of RD. What is needed is an industry partner to co-fund the experimental validation phase — in vitro taste receptor assays followed by in vivo animal preference studies against glucose — and to ensure the research is guided by commercial value, not purely academic interest.
We are looking to connect with a partner who brings formulation expertise, regulatory insight, and a commercial roadmap that transforms a promising compound library into a market-ready product. The sponsoring partner will receive an exclusive option to license the resulting intellectual property, including validated hit compounds and all associated patent rights.
| Ideal Partners ▸ Global food & beverage manufacturers with active sugar-reduction pipelines ▸ Flavor & fragrance companies (Givaudan, IFF, Symrise, Firmenich) ▸ Nutraceutical companies — functional food and dietary supplement formulations ▸ Pharmaceutical companies — taste-masking for oral drug delivery ▸ Biotech / FoodTech startups seeking proprietary sweetener IP | Market Opportunity ▸ $29B+ artificial sweetener market in 2026, growing to $40B+ by 2030¹ ▸ $580B global low-calorie food & beverage sector driving demand for clean-taste ingredients⁶ ▸ $500B+ nutraceuticals market where taste is a primary adoption barrier⁷ ▸ $15B+ pharma taste-masking excipient market with unmet need for novel compounds⁸ ▸ First-to-patent advantage in a novel, unexplored chemical space |
Patents & Publications
Patent Status: Patent application in preparation. The technology has not been publicly disclosed. Yissum is filing for IP protection ahead of a planned peer-reviewed journal publication and scientific conference presentation, expected within 6 months.
| Contact in Yissum Ilya Pittel, Ph.D. | VP of Business Development AgTech, FoodTech, Veterinary & Environmental Sciences Yissum, The Technology Transfer Company of The Hebrew University of Jerusalem T +972.2.658.6693 | M +972.54.782.2354 Hi-Tech Park, Edmond J. Safra Campus, Givat Ram |
References & Market Sources
[1] Grand View Research. Artificial Sweeteners Market Size & Forecast, 2024. https://www.grandviewresearch.com/industry-analysis/artificial-sweeteners-market
[2] Statista. Food & Beverage Industry R&D Spending Worldwide, 2024. https://www.statista.com/statistics/1315614/food-beverage-industry-rd-spending-worldwide/
[3] Mintel Group. Sugar Reduction in Food & Beverages: Market Outlook, 2023. https://www.mintel.com/food-and-drink/sugar-reduction/
[4] Mayo Clinic. Artificial Sweeteners — Health Considerations, 2023. https://www.mayoclinic.org/healthy-lifestyle/nutrition-and-healthy-eating/in-depth/artificial-sweeteners/art-20046936
[5] U.S. FDA. High-Intensity Sweeteners, 2023. https://www.fda.gov/food/food-additives-petitions/high-intensity-sweeteners, https://www.fda.gov/food/food-additives-petitions/aspartame-and-other-sweeteners-food
[6] MarketsandMarkets. Low-Calorie Sweetener Market, 2024. https://www.marketsandmarkets.com/Market-Reports/low-calorie-sweeteners-market-1165.html
[7] Allied Market Research. Nutraceuticals Market Size & Projections, 2023. https://www.alliedmarketresearch.com/nutraceuticals-market
[8] Grand View Research. Pharmaceutical Excipients Market, 2024. https://www.grandviewresearch.com/industry-analysis/pharmaceutical-excipients-market
* EC50 values for novel candidates are pending experimental validation (next research phase).
