Open Source · Research Prototype · Molecule-Aware

A Research Engine For
CNS Delivery Exploration

CEREBRO-X scores candidate drug-delivery-system formulations against an in-house 62-criterion rubric, with a real DNN blood-brain-barrier classifier and a genetic-algorithm formulation optimizer — every computed value carries its method, tier, and citation.

Explore Results ↓ ▶ Watch Demo
98%
CNS drugs fail
to cross the BBB†
62
In-house rubric criteria
per candidate DDS
6
Drugs explored
across 3 molecule classes
$0
Fully open source
MIT licensed

† Pardridge WM (2005) NeuroRx 2:3 — a real, widely-cited literature figure, not a CEREBRO-X result.

Scroll to explore
The Challenge
98%

Fail to cross the
Blood-Brain Barrier

Over 98% of small molecules and ~100% of large biologics fail to reach the CNS in therapeutic concentrations. Researchers spend months in wet-lab trial-and-error when the answer could be computed in minutes.

  • Efflux pumps (P-gp, BCRP) and enzymatic degradation actively reject most candidate nanocarriers at the BBB
  • Protein corona formation on the carrier surface can silently change biodistribution between in-vitro and in-vivo conditions
  • Scale-up from lab to manufacturing introduces shear stress that can collapse fragile nanoparticle architectures
  • Drug-DDS compatibility is rarely assessed computationally — leading to encapsulation failures discovered only after synthesis
  • Glymphatic (CSF) clearance removes brain-resident carriers over time, limiting effective CNS residence
The Process

How CEREBRO-X Works

Physics-informed computation from molecular input to ranked decision report. No wet lab. No waiting. Just answers.

1

Submit Your Molecule

Provide your drug's SMILES/FASTA sequence and specify your candidate DDS formulations. Download our structured Excel template for precise input.

2

62-Principle Analysis

The engine evaluates every candidate against 62 computational principles: BBB physics, PBPK kinetics, protein corona, shelf-life, manufacturability, and more. All driven by real equations.

3

Ranked Decision Report

Receive a comprehensive, publication-ready package: interactive HTML dashboards, PDF reports, animated PBPK videos, a ranked DDS leaderboard, and (on request) a Pre-IND regulatory framework.

Live Case

Same Disease. Different Carriers.

CEREBRO-X ranks candidate DDS formulations against a 62-criterion internal rubric (Class A, surrogate) and then re-checks the winning candidate with real physics — PBPK ODE simulation, DLVO stability, stress-scenario modeling (Class B). The verdicts below are the engine's actual output, unedited — including the cases where deep validation disagrees with the surrogate score.

Drug Class Key Properties Top-1 DDS Class A Score Class B Deep Validation
Donepezil Small Molecule LogP 4.77 · MW 379.5 Da RVG29-PLGA-NP 80.7/100 EXCELLENT PASSED 20/28 (71.4%)
Rivastigmine Small Molecule LogP 2.30 · MW 250.3 Da RVG29-PLGA-NP 83.2/100 EXCELLENT PASSED 20/28 (71.4%)
Galantamine Small Molecule LogP 1.83 · MW 287.4 Da Mannose-Liposome 76.3/100 GOOD MARGINAL 16/28 (57.1%)
Lecanemab Biologic (mAb) ~148 kDa · IgG1 Tf-PEG-Liposome 63.7/100 ACCEPTABLE FAILED 12/28 (42.9%)
Temozolomide Small Molecule LogP −0.96 · MW 194.1 Da RVG29-PLGA-NP 80.8/100 EXCELLENT MARGINAL 19/28 (67.9%)
Nusinersen ASO ~7.1 kDa · Phosphorothioate ApoE-LNP 76.3/100 GOOD FAILED 12/28 (42.9%)
Key Insight: Six drugs, three molecule classes, six different Top-1 carriers — chosen from molecular descriptors alone, no manual tuning. But the honest finding is in the last column: three of six candidates that score well on the fast surrogate rubric (Class A) do not hold up under deep physics re-validation (Class B) at the 70% threshold. That gap is exactly why CEREBRO-X reports both numbers instead of one polished score.
Live Results

Explore Sample Outputs

Real pipeline output for six drugs across three molecular classes. Each card links to a live interactive dashboard and a downloadable publication-ready PDF report.

Small Molecule · BCS I · Alzheimer's
EXCELLENT
Donepezil
AChEI · First-line therapy for mild–severe AD
81
Composite (Class A)
80.7 / 100
LogP 4.77 MW 379.5 Da Class B PASSED 20/28
Optimal DDS
RVG29-PLGA Nanoparticle
View Dashboard ↗ PDF Report
Small Molecule · BCS I · Alzheimer's / PDD
EXCELLENT
Rivastigmine
AChEI · AD and Parkinson's disease dementia
83
Composite (Class A)
83.2 / 100
LogP 2.30 MW 250.3 Da Class B PASSED 20/28
Optimal DDS
RVG29-PLGA Nanoparticle
View Dashboard ↗ PDF Report
Small Molecule · BCS III · Alzheimer's
GOOD
Galantamine
AChEI + nicotinic modulator · Mild–moderate AD
76
Composite (Class A)
76.3 / 100
LogP 1.83 MW 287.4 Da Class B MARGINAL 16/28
Optimal DDS
Mannose-Functionalized Liposome
View Dashboard ↗ PDF Report
Biologic · Monoclonal Antibody · Alzheimer's
ACCEPTABLE
Lecanemab
Anti-Aβ immunotherapy · FDA-approved 2023
64
Composite (Class A)
63.7 / 100
MW ~148 kDa Type IgG1 mAb Class B FAILED 12/28
Optimal DDS
Transferrin-PEG-Liposome (Tf-PEG)
View Dashboard ↗ PDF Report
Small Molecule · BCS I · Glioblastoma
EXCELLENT
Temozolomide
DNA-alkylating agent · GBM standard-of-care
81
Composite (Class A)
80.8 / 100
LogP −0.96 MW 194.1 Da Class B MARGINAL 19/28
Optimal DDS
RVG29-PLGA Nanoparticle
View Dashboard ↗ PDF Report
ASO · Nucleic Acid · Spinal Muscular Atrophy
GOOD
Nusinersen
Antisense oligonucleotide · SMN2 splice modifier
76
Composite (Class A)
76.3 / 100
MW ~7.1 kDa Type Phosphorothioate Class B FAILED 12/28
Optimal DDS
ApoE-Lipid Nanoparticle (LNP)
View Dashboard ↗ PDF Report
Interactive Walkthrough

Experience the Engine

A cinematic presentation of how CEREBRO-X thinks — from BBB physics to ranked output. Navigate at your own pace. All 62 principles explained with real methods and references.

github.com/mohamedtalaat-gif/CEREBRO-X Open fullscreen ↗
The Framework

Powered by 62 Validation Principles

Every candidate DDS is evaluated against 62 computational principles. Three pipeline classes, eight CNS pharma dimensions, one unified composite score.

CNS Delivery

BBB transcytosis, tight junction dynamics, efflux pump interactions, receptor-mediated crossing

Release Kinetics

Dissolution profiles, sustained vs burst release, API-DDS compatibility scoring

Stability

Protein corona, polymorph transformation, shelf-life under Arrhenius kinetics

Safety

hERG, complement activation, microglial neuroinflammation, ADMET off-target toxicity

Glymphatic

CSF clearance trap, Stokes-Einstein diffusion, sleep-cycle-dependent drainage

Manufacturing

Scale-up shear stress, lyophilisation cycles, sterilisability, QbD compliance

Drug-DDS Fit

Ionisation state, encapsulation thermodynamics, ligand-receptor docking (AutoDock Vina)

Translational

Pre-IND framework, FTO & patentability scoring, FDA 21 CFR Part 11 audit trail

Hover any principle to see its class and method.

56 Surrogate (A) 1 Deep Physics (B) 5 Translational (C)
Open Source

Get Involved

CEREBRO-X is a free, open-source research prototype — no fees, no conditions, no required co-authorship. Clone it, run it, file issues, or open a pull request. If it helps your work, a citation (see the repo's CITATION section) is appreciated but never required.

View on GitHub → Read the Engineering Audit Email / Collaborate