UAE Unveils First Fully AI-Discovered Drug Candidate ISM0387 in Under 12 Months

The Dawn of AI-Driven Pharmaceuticals in the UAE

The United Arab Emirates has marked a pivotal moment in its healthcare innovation journey with the announcement of ISM0387, its first fully AI-discovered and developed drug candidate. Unveiled by the Emirates Drug Establishment (EDE) on April 23, 2026, this breakthrough represents a leap forward in accelerating drug discovery for hard-to-treat cancers. Developed entirely within the UAE by a team at Insilico Medicine's research center in Masdar City, Abu Dhabi, the process—from initial molecule design to preclinical candidate nomination—took less than 12 months. This stands in stark contrast to the traditional decade-long timelines that often exceed $1 billion in costs.

ISM0387 targets the PRMT5 enzyme, a protein arginine methyltransferase critical in cellular processes but overexpressed in many cancers. By exploiting synthetic lethality—a strategy where inhibiting PRMT5 proves fatal specifically to cancer cells lacking the MTAP gene—the compound selectively attacks tumors while sparing healthy tissue. MTAP deletions occur in 40-50% of glioblastoma multiforme (GBM) cases, the most aggressive primary brain cancer, which boasts a grim 70% recurrence rate and claims over 200,000 lives globally each year.

This achievement underscores the UAE's strategic pivot toward a 'Falcon Economy,' emphasizing technology-driven growth in high-value sectors like biotechnology. Through partnerships under the Abu Dhabi Investment Office's (ADIO) Health, Endurance, Longevity, and Medicine (HELM) cluster, the nation is fostering local R&D capabilities to secure pharmaceutical sovereignty.

Understanding ISM0387: Mechanism and Promise

Protein arginine methyltransferase 5 (PRMT5) methylates arginine residues on proteins, influencing gene expression, RNA splicing, and signal transduction. In MTAP-deficient cells—common in GBM, pancreatic, and lung cancers—PRMT5 inhibition disrupts methionine salvage pathways, leading to toxic metabolite accumulation and cell death. ISM0387, an MTA-cooperative PRMT5 inhibitor, enhances this effect, showing superior selectivity over first-generation inhibitors plagued by toxicity.

Preclinical data reveal ISM0387's prowess: it penetrates the blood-brain barrier effectively across species, boasts a favorable pharmacokinetic profile (low molecular weight, optimal polar surface area, balanced lipophilicity), and demonstrates dose-dependent tumor suppression. In orthotopic GBM mouse models, daily oral dosing at 30 mg/kg over 20 days halted tumor progression, unlike vehicle controls where tumors advanced rapidly. In vitro, it exhibited potent activity across MTAP-deleted cell lines with a clean safety margin.

• Improved in vitro potency and selectivity compared to prior PRMT5 inhibitors.
• Enhanced brain exposure for CNS targeting.
• Robust efficacy in relevant disease models.
• Favorable ADMET properties for human translation.

Glioblastoma, accounting for 48% of primary malignant brain tumors, remains notoriously resistant to therapy, with median survival under 15 months post-diagnosis. ISM0387's profile positions it as a potential game-changer, especially for recurrent cases.

The Power of Chemistry42: Insilico's AI Engine

At the heart of this feat is Chemistry42, Insilico Medicine's proprietary generative AI suite comprising over 40 models trained on vast biological, chemical, and clinical datasets. Unlike rule-based screening, Chemistry42 employs reinforcement learning and generative adversarial networks to de novo design molecules, predict binding affinities, optimize synthesizability, and forecast ADMET outcomes.

The workflow for ISM0387 began with AI-driven target validation for PRMT5 in MTAP-deleted contexts. It then generated 90 novel structures using co-crystal data from known inhibitors, prioritizing CNS attributes like LogP, pKa, and brain-plasma ratios. Virtual screening narrowed candidates, followed by rapid synthesis and testing by the four-person UAE team (two computational chemists, one medicinal chemist, one biologist). Lead optimization refined four top hits into ISM0387 in six months—a phase traditionally spanning years.

This end-to-end AI integration slashed timelines: target ID (weeks vs. months), hit generation (<30 days), and PCC nomination (under 12 months total). Insilico, with 40+ prior nominations and three Phase II assets, credits Pharma.AI for 60-200x efficiency gains.

From Concept to Candidate: A Compressed Timeline

Traditional drug discovery follows a linear path: target validation (1-2 years), hit identification (2-3 years), lead optimization (3-4 years), preclinical nomination (1 year)—totaling 7-10 years pre-IND at $1B+. ISM0387 inverted this.

The UAE team's small size highlights AI's scalability, enabling boutique operations to rival pharma giants. This model promises democratized innovation, vital for addressing unmet needs in oncology.

UAE's Vision: Building a Biotech Powerhouse

The UAE's Operation 300bn targets AED 300 billion in industrial GDP by 2031, with pharma as a pillar. EDE, established 2022, regulates and innovates, partnering with ADIO's HELM for longevity-focused R&D. Insilico's 2023 Masdar entry, MoU with EDE (Feb 2026), catalyzed this.

Quotes reflect ambition: EDE Chairman Saeed Al Hajeri hailed it as maturing a national model blending research, regulation, investment. Director-General Fatima Al Kaabi emphasized local medicine development redefining security. ADIO's Badr Al-Olama sees scalable export potential. Insilico's Alex Aliper: "UAE evolved from trade to science hub."

Collaborations with MBZUAI, Khalifa University, NYU Abu Dhabi bolster talent pipelines.

Global Context: AI Reshaping Drug Discovery

AI's pharma ascent is global: Insilico's prior IPF candidate (ISM001), Phase IIa-positive; Exscientia's oncology deals; Recursion's neurology pipeline. Yet UAE's fully local execution is unique, signaling emerging markets' rise.

Challenges persist: AI validation, regulatory hurdles, data quality. Successes like ISM0387 validate generative models, potentially halving costs/timelines industry-wide. For GBM, where options dwindle post-TMZ, targeted therapies like PRMT5 inhibitors fill voids.Learn more on PRMT5 inhibitors

Preclinical Promise and Path Forward

ISM0387's data: superior selectivity, BBB penetration, tumor stasis in models. Next: IND filing, Phase I safety in healthy volunteers, then oncology trials. EDE's framework fast-tracks approvals, aiming global patients.

• Safety: Clean in vitro profile.
• Efficacy: Dose-proportional in GBM xenografts.
• Translation: Human-relevant PK.

Risks: Clinical attrition (90% fail), but AI's precision mitigates.

Implications for UAE Healthcare and Economy

Beyond medicine, this spurs jobs (Insilico's 40 specialists), IP generation, manufacturing. UAE's 2026 AI Strategy allocates billions, positioning it against Singapore, Israel in biotech.

Cancer burden: 19,000 new cases yearly; local innovation cuts import reliance (95% drugs imported).

Challenges and Ethical Considerations

AI biases, data privacy, equity access loom. UAE mandates ethical AI, diverse datasets. Validation rigor ensures safety.

Future Outlook: More Breakthroughs Ahead

EDE-Insilico eyes fibrosis, immunology. HELM envisions export hub. Globally, AI could yield 50 new drugs/decade, saving trillions.EDE's full announcement KT coverage

This UAE milestone heralds an era where AI turns data into cures swiftly, benefiting humanity.

Photo by Markus Winkler on Unsplash