Simulaite Report

This methodology demo compares Formulaite Metabolic Support against a market reference formula by asking which modeled active markers reach the right biological sites, at meaningful concentrations, with enough potency and structural support to count. The score is intentionally conservative: a marker-target pair contributes only when activity, exposure, and structural evidence all support the same mechanism.

The workflow combines active-marker dose modeling, Asian-population oral PBPK, target-site concentration-time profiles, IC50/EC50 activity prediction from molecular descriptors and graph neural network models, quantum-chemistry-supported ligand preparation, docking against concrete target pockets, pose de-risking, and ingredient contribution analysis.

Bottom Line: Formulaite Shows Broader Support, Including GLP-1/GIP Release Signaling

• Formulaite Metabolic Support shows the stronger double-gated signal for carb absorption support and GLP-1/GIP release & satiety signaling support.
• Market Reference shows one clear tradeoff advantage: stronger GLP-1/GIP incretin preservation support, mainly through the HCA-DPP4 mechanism.
• The final comparison therefore favors Formulaite for broader modeled metabolic-support coverage.
• The model highlights a practical formulation principle: ingredient names are not enough; dose must be sufficient for target-site exposure that is meaningful relative to predicted potency.

Table 1. Final Double-Gated Modeled Scores

Oral PK outputs provide plasma, gut-local, and target-site concentration-time curves. Activity models estimate targetspecific IC50 or EC50 potency. Docking is then used as a structural de-risking layer rather than as potency by itself.

How the Score Is Built

• First, each extract is converted into modeled active-marker doses from the daily extract dose, standardization percentage, and marker split.
• The PBPK layer estimates how much of each marker reaches the target-relevant site over time. Intracellular, mucosal, lumen, and plasma-facing targets use the appropriate concentration-time profile.
• The activity layer predicts target-specific IC50 or EC50 potency. Engagement is computed directly from C(t)/(C(t)+potency), so exposure below potency contributes little or nothing.
• The docking layer compares each marker against the same target pocket and acts as structural evidence. The pair must clear the docking floor and pose check, then its activity score is scaled by the de-risked docking score.
• Each molecular target contributes to one primary endpoint only. DPP4 counts in GLP-1/GIP preservation, MGAM in carb absorption, and GPR119 in GLP-1/GIP release and satiety signaling.
• A zero in this modeled lane means no marker-target pair survived all evidence gates for that endpoint; it does not mean the biology is impossible.

The PK simulation uses the declared daily serving regimen as scheduled oral dose events rather than collapsing the day into one bolus dose. This matters because target-site exposure is time-dependent.

Formulaite Metabolic Support

Market Reference Formula

Fenugreek, gymnema, Triphala, bitter melon, and black

Garcinia cambogia, Terminalia chebula, Fenugreek,

pepper.

Gymnema sylvestre, and Guggul resin. Unknown standardizations are modeled using explicit comparator assumptions.

Modeled marker dose is based on extract dose, active percentage, and marker split, not full extract mass. Triphala is modeled as 45% tannins/polyphenols.

These modeled marker doses feed the PK layer directly, so low marker dose can limit target-site exposure even before potency or docking evidence is considered.

The PK layer uses a PBPK-style virtual-cohort simulation to estimate plasma, gut-local, and tissue-adjacent exposure for each modeled active marker after oral dosing. The same population and meal context are used for both products.

Virtual Population and PBPK Conditions

The same virtual cohort is used across formula conditions. F% is computed as oral AUC divided by matched IV AUC for the same population cohort.

Table 2A. Formulaite Metabolic Support PK Metrics

Figure 1A. Scheduled median plasma concentration-time curves for modeled active markers in Formulaite Metabolic Support using 1 capsule x 3 daily dose events. Curves are shown on a log concentration scale so lower-exposure markers remain visible.

Table 2B. Market Reference Formula PK Metrics

Figure 1B. Scheduled median plasma concentration-time curves for modeled active markers in Market Reference Formula using 2 capsules x 2 daily dose events, with the same virtual population and meal context.

The headline scores are not based on one generic plasma exposure number. Each target uses the exposure site that matches the biology: DPP4 uses systemic plasma, MGAM uses gut lumen exposure, and GPR119 uses small-intestine mucosa interstitial exposure.

Engagement is the PK/PD bridge: engagement(t) = C(t) / (C(t) + IC50 or EC50). If target-site exposure is far below predicted potency, engagement stays near zero even when docking looks acceptable. Passing pairs are then scaled by the de-risked docking weight.

Table 3. All Modeled Pair Contributions, Ranked by Score

This table includes every nonzero modeled pair that passed the activity, target-site exposure, and same-pair docking gates, ranked by final pair contribution.

Failed pairs are excluded from this table. Cmax and AUC are shown in the target-site compartment actually used for the pair score, not necessarily plasma.

Engagement is the average C(t)/(C(t)+potency) over the scoring window.

The tables below show the final double-gated modeled scores. A pair must have target-site exposure, an IC50/EC50 activity estimate, and same-pair docking support that clears the structural floor. Passing pairs are scaled by the de-risked docking score, so weaker structural support contributes less even after passing the gate.

Why Some Familiar Ingredients Contribute Little Here

• Ingredient contribution is driven by modeled marker dose, target-site exposure, predicted IC50/EC50 potency, and same-pair docking support together; high extract mass alone does not guarantee a modeled score.
• The same ingredient marker can contribute in one formula but drop to zero in another if the lower-dose formula does not reach target-site concentrations that are meaningful relative to predicted potency, as seen for some lower-dosed Market Reference ingredients.

Carb absorption support

Gut-local carbohydrate enzyme biology, driven here by MGAM. A higher score means stronger modeled modeled support

for slowing carbohydrate breakdown or absorption at the intestinal target site.

Formulaite contributors

Market Reference contributors

MGAM Mechanism Spotlight

The strongest carb absorption support signal in Formulaite Metabolic Support is chebulinic acid from Triphala on MGAM.

The panel pairs a quantum electrostatic surface with a locked-view docking comparison against the MGAM acarbose reference pocket.

Figure 2. Chebulinic acid is modeled from the Triphala extract fraction standardized to 45% tannins/polyphenols.

GLP-1/GIP incretin preservation support

DPP4-linked preservation of GLP-1/GIP incretin signaling. A higher score means stronger modeled modeled support for

reducing incretin breakdown after release.

Formulaite contributors

Market Reference contributors

GLP-1/GIP release & satiety signaling support

GPR119-linked gut-hormone release and satiety signaling. A higher score means stronger modeled modeled support for GLP-1/GIP release-adjacent gut hormone signaling.

Formulaite contributors

Market Reference contributors

Tracked but Not Supported in This Run

The workflow also tracked additional mechanistic areas. They are reported as zero in the modeled lane because the available marker-target evidence did not pass the full activity, target-site exposure, and same-pair docking gate.

The computational benchmark is best used to prioritize a small set of confirmatory assays. These follow-ups focus on the modeled mechanisms that survived the IC50/EC50, target-site exposure, and docking de-risking workflow.