Comprehensive analysis of commercial fragment libraries.

Description of the files

Fragments from libraries:

• allFragments_annotations.sdf: SDF file containing the 3D conformation of all unique 512284 fragments.
  • Descriptors present on the file: Library type (type), Library size (size) and the number of times the fragment is present in the ensemble of libraries (Frequency).
• unique_nonbiased_MW_logP_PBF.sdf: SDF file containing the ensemble of unique fragments from the non-chemically biased libraries.
  • Descriptors present on the file: Canonical Smiles (Canonical_Smiles), Molecular Weight (MW), Partition logarithm (ALogP) and Plane of best fit value (PBF).
• unique_nonbiased_principal_descriptors.tsv: TSV file containing the ensemble of unique fragments from the non-chemically biased libraries with the descriptors that were explicitely used in the article.
  • Descriptors present on the file: Canonical Smiles (Canonical_Smiles), molecular weight (MW), partition logarithm (ALogP) and plane of best fit value (PBF), number of heavy atoms (HAC), number of hydrogen bond acceptors and donors (HBA and HBD), number of rotatable bonds (NRot), number of violations of the rule of three (Ro3_Vio) and quantitative estimate of drug-likeness (QED).
• unique_nonbiased_descriptors.tsv: TSV file containing the ensemble of unique fragments from the non-chemically biased libraries with all the calculated descriptors.

GTM model:

• training_set_GTM.sdf: SDF file containing the set of 6017 fragments used for training the GTM model.
• Model_IIAB2-4_cycle_Freq_05_m301.xml: xml file of the generated model allowing the generation of GTM landscapes.

Scaffold data:

• scaffold_data.tsv: TSV file containing the scaffold data. For an easy exploration of the file please refer to: https://gtmfrag.drugdesign.unistra.fr/