ALCHIMIA: Generated Molecules for CB2R and S1R Receptor Targets

A 1.4 MB dataset containing molecules generated by the ALCHIMIA framework for computational drug discovery. The framework, developed by Domenico Alberga, combines reinforcement learning and a genetic algorithm based on 33 medicinal chemistry-inspired transformations. It was applied to the human Cannabinoid Receptor 2 (CB2R) and Sigma nonopioid intracellular Receptor 1 (S1R) for hit identification, lead optimization, and dual modulator design.

Use Cases

• Benchmarking generative molecular design algorithms based on the described synthetic accessibility (SA) and drug-likeness (QED) scores.
• Training or validating policy networks for reinforcement learning in chemistry based on sequences of molecular transformations.
• Analyzing chemical lineages and population diversity in genetic algorithm-based optimization for drug discovery.
• Studying multi-objective optimization strategies for balancing target affinity, synthetic accessibility, and drug-likeness.

Strengths

• Dataset is based on a defined vocabulary of 33 medicinal chemistry-inspired molecular transformations.
• Molecules were generated for two specific, pharmacologically relevant targets: CB2R and S1R.
• Framework performance was evaluated against random baselines and selected de novo design methods, as noted in the description.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• Description metadata is limited; actual data quality requires manual inspection after download.

Provenance

Source

figshare, author Domenico Alberga.

Collection Method

Generated by the ALCHIMIA computational framework combining reinforcement learning and a genetic algorithm.

Freshness

Last updated 2026-04-16 07:29:47; freshness should be verified.