Computational Pharmaceutics

Muhammad Talaat
Cheminformatics & Drug Delivery System Engineering

BPharm, Faculty of Pharmacy — Ain Shams University. I build open-source computational tools at the intersection of cheminformatics, machine learning, and pharmaceutical formulation science, and publish peer-reviewed research on nanomedicine and drug delivery.

BPharm · Ain Shams University RDKit · DeepChem · scikit-learn Cairo, Egypt

01About

I'm a pharmacist and computational researcher working on tools that connect cheminformatics, machine learning, and pharmaceutical formulation engineering — with a focus on CNS drug delivery and nanomedicine. My long-term goal is to help build a coherent, reproducible, open-source ecosystem for computational pharmaceutics: engine that pharmaceutical researchers can actually trust, verify, and build on, not black-box tools with unverifiable claims. I care as much about honest reporting of what a model does and doesn't do as I do about the model itself.

02Research Interests

Cheminformatics

  • Molecular representation & standardization
  • Molecular descriptors & fingerprints
  • QSAR / ADMET modeling

Drug Delivery

  • Nanoparticle & carrier formulation
  • Blood-brain barrier permeability
  • CNS-targeted delivery systems

Machine Learning

  • Molecular property prediction
  • Deep learning for cheminformatics
  • Formulation-parameter optimization

Open Science

  • Reproducible research engine
  • Data curation & quality control
  • Benchmarking & honest reporting

03Projects

CEREBRO-X Research Prototype

Computational pipeline for CNS drug delivery system exploration

An open-source pipeline that scores candidate drug-delivery-system formulations against an in-house, transparently-documented 62-criterion rubric, and includes a real deep-learning blood-brain- barrier permeability classifier (trained on the public BBBP benchmark, ~90% held-out accuracy, RDKit + TensorFlow) plus a genetic-algorithm formulation optimizer. Every computed value carries an explicit tier, method, and citation — including which parts are literature-grounded correlations versus in-house heuristics not yet independently validated.

Python RDKit TensorFlow AutoDock Vina FastAPI Docker
Status, honestly: research-stage, not clinically or experimentally validated, and under active development. The full engineering & scientific-integrity audit is public in the repo. I'd rather you read that than take a marketing description at face value.

View on GitHub →

04Publications

Nano-antioxidants for neurodegenerative disorders: a scoping review

Discover Nano · 2025

Biologically synthesized nanoparticles: barley-mediated silver and gold nanoparticles and caged gold nanoplatform for advanced drug delivery system engineering in medicine

Discover Nano · 2024

Full, up-to-date list on ORCID (0000-0002-9171-437X).

05Get in Touch

Open to research collaborations, PhD/postdoc opportunities, and conversations about open-source computational pharmaceutics tooling.