Data used in the "Generative Molecular Design with Steerable and Granular Synthesizability Control" pre-print

This deposit contains open-source data sets used in the pre-print:<br>Raw PubChem cleaned Saturn pre-training data eMolecules building blocks stock used in the retrosynthesis model (isotopes removed)100 enforced blocks which is a subset of the eMolecules stock - used in the block enforcing experimentsUS (glycine removed) and EU (formic acid, aniline, succinic acid removed) waste blocks used in the waste valorization experiments. Certain waste blocks were removed to show more diversity of blocks and reactions in the synthesis routes. This way, we could keep the objective function fixed and interrogate enforcing other blocks (blocks can be more suited for certain objectives)Some datasets are not deposited here because we cannot release them:<br>Enamine building blocks which can be obtained by request here: https://enamine.net/building-blocks/building-blocks-catalogChemspace's Freedom 4.0 space building blocks and synthons which can be obtained by email request: [e-mail]