Accelerating Drug Discovery with AI

Harness the power of machine learning and computational chemistry to predict drug properties, optimize leads, and reduce time-to-market.

About DisSolved

Dis-Solved is an independent AI-powered platform dedicated to making advanced in silico drug discovery accessible, affordable, and targeted—especially for central nervous system (CNS) therapeutics and safer stimulant design.

Founded and built single-handedly by Nabil Yasini, Dis-Solved started as a personal project on a modest laptop: developing high-performance, interpretable machine learning models that outperform many state-of-the-art benchmarks in key ADMET predictions. What began as a breakthrough blood-brain barrier permeability (BBBP) predictor (achieving 0.92 internal AUC and 0.96 on external validation, with explicit stereoisomer training) has evolved into a modular toolkit tailored for early-stage screening of ADHD and neuropsychiatric drug candidates.

Our Core Focus

  • Specialized predictions for monoamine transporter activity (including kinetics for DAT, NET, SERT)
  • Toxicology screening and safety profiling
  • PK/PD curve simulation for pharmacokinetic modeling
  • Abuse potential scoring—to help design lower-liability stimulants

By leveraging clever data augmentation and modern graph neural networks, we've created tools that are not only accurate and explainable but also beginner-friendly and priced for real-world use—starting from free tiers up to affordable subscriptions.

At Dis-Solved, we believe cutting-edge AI shouldn't be locked behind enterprise paywalls or massive teams. We're here to empower academics, small biotechs, indie researchers, and drug hunters to screen smarter, fail faster, and advance safer CNS therapies.

Whether you're a student exploring computational chemistry or a startup optimizing leads, Dis-Solved is your practical partner in dissolving the toughest challenges in drug discovery.

0.96 AUC

External Validation

Stereo-Aware

GNN Architecture

Free Tier

Available

Indie Built

By Nabil Yasini

Blood-Brain Barrier Permeability (BBBP)

State-of-the-art GNN model achieving AUC 0.9612 on external validation. Predict whether compounds can cross the blood-brain barrier for CNS drug development.

Insilico Drug Discovery Toolkit

A comprehensive suite of machine learning models for predicting key drug properties and potential liabilities early in the discovery pipeline.

MAT Transporter Abuse Potential Cardiotoxicity Metabolism

PK/PD Simulations

Interactive pharmacokinetic and pharmacodynamic modeling tools to simulate drug behavior, optimize dosing regimens, and predict therapeutic outcomes.

Coming Soon!

We're building interactive PK/PD simulation tools including one & two-compartment models, population PK, and dose optimization capabilities.

Compartment Models Parameter Estimation Population PK Dose Optimization

Custom Services

Need specialized models or consulting? Our team of computational chemists and data scientists can help accelerate your drug discovery programs.

Bespoke Model Development

Custom ML/AI models trained on your proprietary data for specific endpoints, targets, or compound classes.

  • Target-specific QSAR models
  • Virtual screening workflows
  • De novo molecular generation

Consulting & Training

Expert guidance on implementing AI-driven drug discovery strategies and training your team on best practices.

  • Workflow optimization
  • Team workshops & seminars
  • Technology evaluation

Platform Integration

Deploy DisSolved tools in your infrastructure with API access, custom dashboards, and enterprise support.

  • REST API deployment
  • Cloud & on-premise options
  • ELN/LIMS integration

Get in Touch

Have questions about our platform or interested in collaboration? We'd love to hear from you.