AIDDISON Explorer

Overview

AIDDISON Explorer supports molecular design and candidate prioritization in drug-discovery workflows. The supplied product-page evidence describes a hosted platform that accepts a target profile and molecular design constraints, then generates structures with reinforcement-learning models based on REINVENT 4.0. Its advertised applications are hit identification, scaffold-focused hit expansion and lead optimization. The workflow focuses on generating and prioritizing candidates for subsequent evaluation.

Generated candidates are scored across potency, physicochemical and developability properties, and ADMET criteria before being ranked. ADMET predictions include confidence and explainability information, which can help researchers assess the basis and uncertainty of a prediction. Synthetic accessibility scoring and routes obtained through the SYNTHIA API add feasibility considerations, including route step count and building-block availability. These outputs support candidate selection but do not establish experimental activity, safety or successful synthesis.

Ranked molecules and predicted properties can be exported in CSV, SDF and RD formats for downstream docking, synthesis planning and experiments. Explorer therefore fits between a project's design requirements and subsequent computational or laboratory assessment. Access begins with requesting a vendor demonstration and discussing project needs; the supplied page does not provide public local-install commands or an unrestricted public API. Its capability descriptions are vendor claims, not independent validation, and candidate rankings and model predictions require experimental follow-up.

Key Features

• Generates molecular structures from a target profile and design constraints using reinforcement-learning models based on REINVENT 4.0.
• Scores and ranks candidates across potency, physicochemical and developability properties, and ADMET criteria.
• Provides confidence and explainability information with ADMET predictions.
• Combines synthetic accessibility scoring with SYNTHIA API routes that consider feasibility, step count and building-block availability.
• Exports ranked molecules and predicted properties in CSV, SDF and RD formats for downstream workflows.

Use Cases

• Intended evaluation for hit identification: assess whether generated and ranked candidates provide a useful shortlist for subsequent docking and experimental testing.
• Intended evaluation for scaffold-focused hit expansion: explore candidates under project-defined molecular constraints and compare their predicted properties before follow-up.
• Intended evaluation for lead optimization: assess trade-offs among predicted potency, developability, ADMET and synthetic feasibility when choosing molecules for synthesis planning.

How to Use

1. Review the official AIDDISON Explorer page and request a vendor demo. Discuss the intended project workflow and access arrangements; no public local-install procedure is supplied.
2. Prepare a target profile and molecular design constraints. Use the demo discussion to clarify how these inputs would be represented for your project rather than assuming an undocumented input schema.
3. Evaluate generated candidates against the platform's potency, physicochemical/developability and ADMET scores. Examine the confidence and explainability information accompanying ADMET predictions.
4. Review synthetic accessibility scores and SYNTHIA API route information, including feasibility, step count and building-block availability, when selecting candidates for further investigation.
5. Export ranked molecules and predicted properties in an appropriate supported format—CSV, SDF or RD—for downstream docking, synthesis planning and experiments. Treat the rankings as prioritization aids and arrange experimental follow-up rather than treating predictions as confirmed results.