Novus Nexum Labs — Research Paper

Computational Platform for Peptide Compounding & Clinical Operations

Schema-driven pharmacokinetic models, automated safety checks, and cryptographic audit trails — purpose-built for the daily clinical, compounding, and lab workflows that peptide pharmacies actually run. Third-party validation in progress.

0
Deterministic Tools
0
Namespaces
0
Pre-flight Checks
HMAC
Signed Certificates
Scroll

The Peptide Compounding Landscape Is at an Inflection Point

A $53 billion peptide therapeutics market collides with the most aggressive FDA enforcement cycle in compounding history. The clinics, pharmacies, and labs that survive will be the ones whose compounding operations and clinical documentation are computationally auditable from formulation to patient.

$53B+

Peptide Therapeutics Market

Global market approaching $70B by 2031. Metabolic, oncology, and personalized peptide therapies are driving unprecedented demand through compounding channels.

503A/B

Regulatory Reckoning

FDA has resolved national GLP-1 shortages and ended enforcement discretion. Compounders operating on shortage exemptions face immediate compliance requirements.

30+

Warning Letters (2025)

Telehealth platforms, 503B facilities, and even contract labs cited for potency failures, sterility gaps, and misleading marketing of compounded peptides.

Zero

No Computational Tools

Enterprise PBPK platforms serve big pharma at enterprise cost. Compounders and clinics rely on spreadsheets, manual calculations, and undocumented shortcuts.

Part II

The Operations Gap

Between the pharmacokinetic science a compounder should document and the tools they actually have access to for daily clinical and lab operations, there is a dangerous void.

Enterprise pharmacokinetic platforms — Simcyp, GastroPlus, NONMEM — were built for billion-dollar drug development programs. They carry six-figure price tags, require specialist operators, and produce outputs formatted for NDA submissions, not pharmacy SOPs. They were never designed for the compounding pharmacist checking a dosing rationale, the lab tech verifying unit consistency, or the QA lead assembling inspection evidence.

On the other end, peptide clinics and 503A/503B compounders face daily operational questions — What is the accumulation ratio at steady state? Are we within the therapeutic window for a renally-impaired patient? Does this formulation maintain dimensional consistency across the compounding record? — but answer them with Excel, institutional memory, and hope.

FDA's post-shortage enforcement posture makes this untenable. Warning letters increasingly cite unsupported dosing rationale, subpotent releases, and inadequate out-of-specification handling. The question is no longer whether you have the right answer — it's whether you can prove it.

"The issue is not that compounders lack pharmacological knowledge. The issue is that their knowledge leaves no trace."

Today's Reality

  • Dosing calculations in unversioned spreadsheets
  • No dimensional analysis — unit errors pass silently
  • BUD estimation based on convention, not kinetic models
  • Drug interaction checks done by memory
  • No computational audit trail for inspectors
  • Sensitivity analysis never performed

With Functo

  • Schema-driven PK models with pre-flight checks
  • Automatic unit propagation and dimensional verification
  • Accumulation ratios and steady-state bounds computed
  • Compound-risk analysis with symbolic perturbation
  • HMAC-signed Merkle DAG certificates per computation
  • One-click sensitivity tornado for every parameter
Part III

The Functo Platform

A computational engine built for clinical compounding, lab operations, and AI agents alike — 717 deterministic tools across 29 namespaces for dosing verification, safety profiling, formulation QA, and generating inspector-ready evidence. Every result hashed, every chain certifiable. Third-party validation in progress.

🔍
Classify
Route to the right model template
Validate
6 pre-flight checks, auto
Safety Check
Bounds, accumulation, risk
🔒
Certify
HMAC-SHA256 Merkle DAG
📈
Visualize
Publication-quality output
01 — MODEL ENGINE

Schema-Driven PK/PD Models

Define model templates with typed parameters, bind patient-specific values, and let six automated pre-flight checks — completeness, domain, semantic, dimensional, assumption, and solvability — validate before any computation runs.

503A/503B 1-compartment Multi-compartment Renal Adjustment
02 — SAFETY CHECKS

Automated Post-Flight Analysis

Five post-flight safety checks evaluate every instantiated model: constraint bounds with margins, steady-state accumulation ratios, compound-risk perturbation for DDIs, local sensitivity coefficients, and kinetic regime discrimination.

Bounds Accumulation Compound Risk Sensitivity Discriminant
03 — DIMENSIONAL ANALYSIS

Unit Safety That Catches What Humans Miss

Pint-backed unit propagation through every expression. Buckingham Pi theorem for scale-invariant summaries. Automatic dimensional verification flags mg/mL vs. mg/L mismatches before they reach a patient.

Pint Engine Buckingham Π Auto-propagation
04 — CERTIFICATION

Cryptographic Audit Trails

Every computation is a node in a Merkle DAG. At session end, an HMAC-SHA256 signed certificate captures the full provenance chain — who computed what, with which inputs, producing which outputs. Offline-verifiable. Inspector-ready.

Merkle DAG HMAC-SHA256 Offline Verify Audit Export
05 — VISUALIZATION

Publication-Quality Pharma Charts

53 visualization tools including pharma-specific outputs: VPC plots, goodness-of-fit panels, forest plots with confidence intervals, spaghetti fits, Kaplan-Meier curves, dose-response, and sensitivity tornados.

VPC GOF Panel Forest Plot Tornado HTML/PDF Export
06 — PEPTIDE INTELLIGENCE

Structure-Aware Peptide Analytics

33 cheminformatics tools for peptide-specific lab workflows: molecular descriptors and fingerprint similarity for identity verification, QSAR modeling for property prediction, and ADMET-style profiling for formulation assessment. Same engine extends into novel compound research and clinical trial support as your operation scales — a capability we are actively building toward.

RDKit QSAR Fingerprints ADMET Identity Verification Trials Roadmap

Compounding Pharmacy Pain Points

Ranked by operational impact — 2025 industry survey data
Regulatory burden
94%
BUD/stability gaps
87%
Dosing doc burden
82%
Unit/calc errors
76%
DDI checking
71%
Audit readiness
68%

Functo Coverage by Regulatory Domain

Platform capabilities mapped to compliance requirements
90% Coverage
70% — Direct computational coverage (PK models, units, bounds, audit)
20% — Analytical support (viz, reporting, sensitivity)
10% — Physical/lab processes (out of scope)
Part IV

In Practice: A Peptide Formulation Workflow

Consider a 503B outsourcing facility validating a BPC-157 subcutaneous peptide formulation for a clinic network. The pharmacist needs to demonstrate therapeutic window compliance, accumulation safety, and dimensional consistency — with evidence that survives an FDA inspection.

Pipeline: Subcutaneous Peptide Model

Five steps. Full audit trail. Zero spreadsheets.

STEP 01

Classify & Route

Functo's schema classifier identifies the elimination pathway and recommends the appropriate 1-compartment SC model template.

AUTO-ROUTED
STEP 02

Instantiate & Validate

Bind patient parameters (weight, eGFR, dose). Six pre-flight checks run automatically: completeness, domain, semantic, dimensional, assumptions, solvability.

6/6 PASSED
STEP 03

Safety Bounds

Post-flight bounds check evaluates Cmax, Css, and trough concentrations against the therapeutic window. Margin of safety computed per constraint.

WITHIN BOUNDS
STEP 04

Accumulation & Sensitivity

Accumulation ratio R calculated at the dosing interval. Sensitivity tornado identifies which parameters most affect steady-state exposure.

R = 1.34
STEP 05

Certify & Export

HMAC-SHA256 certificate generated over the computation DAG. Publication-quality report exported with provenance chain intact.

CERTIFIED

Computation Provenance Chain

✔ Verified
Schema Define
a3f8...c41d
Instantiate
7b2e...9f03
Bounds Check
d91c...4a77
Accumulation
e5f1...b208
Merkle Root
f2a9c7...3d18e4
Part V

Compliance Coverage

Functo capabilities mapped directly to regulatory requirements facing peptide compounders and clinics.

Regulatory Requirement Framework Functo Capability Covered
Dosing rationale documentation 503A / 503B Schema-driven PK models with pre-flight checks
Dimensional consistency USP <797> Pint unit propagation + Buckingham Π analysis
Steady-state safety 503B cGMP Accumulation ratio + bounds check with safety margins
Drug interaction analysis 503A / 503B Compound-risk perturbation with symbolic re-evaluation
Parameter sensitivity cGMP / QA Local sensitivity coefficients + tornado visualization
Computation audit trail 21 CFR Part 11 HMAC-SHA256 signed Merkle DAG certificates
Reproducible evidence FDA Inspection Session checkpoints, run comparison, provenance export
Formulation optimization USP <797> Peptide descriptors, QSAR, stability-indicating analytics
Peptide identity & characterization Lab / QC Molecular descriptors, fingerprint similarity, QSAR property checks
Publication-ready reporting QA / Legal 53 pharma visualization tools + HTML/PDF export

Why Now

Four converging forces make computational infrastructure not a luxury, but a survival requirement for peptide compounders.

2025

Shortage Discretion Ends

FDA resolved semaglutide and tirzepatide shortages. Enforcement discretion for compounding "essentially a copy" has expired. Compounders must now justify every formulation on its own scientific merits.

2025–26

Warning Letter Escalation

Simultaneous enforcement from FDA inspections, state boards, payer audits, and innovator litigation (Novo Nordisk, Eli Lilly) narrows the viable operating corridor for undocumented operations.

2026

Market Bifurcation

The compounding market is splitting: high-compliance, QA-invested operations win institutional contracts, while undocumented "peptide mills" face existential legal and regulatory risk.

2026+

From Operations to Trials

Functo is built as an MCP server — AI-callable, deterministic, and auditable from day one. The same platform powering your compounding operations today extends into novel compound research and clinical trial support as we build that capability next.

"In a post-shortage regulatory environment, the peptide compounder's most valuable asset is not their cleanroom — it's the provenance chain connecting their compounding operations to their patient."

— nnLabs, 2026

Functo does not replace the pharmacist's judgment. It captures it — in a form that is reproducible, verifiable, and defensible. Every model schema, every safety check, every sensitivity analysis becomes part of an immutable computation record that answers the only question an inspector will ever ask:

"Show me how you know this is safe."

And as your operation grows, the same infrastructure extends. We are actively building toward novel compound research and clinical trial support — so the platform that runs your pharmacy today becomes the platform that powers your pipeline tomorrow.

Ready to Upgrade Your Compounding Operations?

Schedule a demonstration of the Functo platform. See your formulations modeled, your dosing verified, and your evidence certified — all in one session.

Request a Demo ➔ Back to Top ↑