SEARCH

SEARCH BY CITATION

FilenameFormatSizeDescription
cbdd12155-sup-0001-Suppinfo.docWord document184K

Appendix S1. Custom-built python script to randomly split all targets into training and test sets.

Appendix S2. List of ChEMBL ID-s for targets of Cutoff-30 subsets.

Appendix S3. Custom-built python script to perform advanced compound selection based on training targets.

Table S1. Test Target coverages measured for subsets of compounds generated by fingerprint-based compound selection from Cutoff-30.

Table S2. Average test target coverages measured for subsets generated by Method 2 (bioactivity-based compound selection using Script 1) from Cutoff-30 dataset.

Table S3. Average test target coverages for Cutoff-30 subsets generated by Method 3 (bioactivity-based compound selection using Script 2) are shown above in addition to the percentage improvement compared to Method 2 (using Script 1).

Table S4. Demonstrates direct comparison of fingerprint-based compound selection and the optimized bioactivity-based compound selection method.

Figure S1. Scheme for generation of MOLPRINT2D fingerprints.

Figure S2. Test target coverage for 0.5–10% subsets generated by fingerprint-based compound selection from Cutoff-30 and Cutoff-40 datasets.

Figure S3. Promiscuity of Cutoff-30 compounds selected by fingerprint-based and bioactivity-based (using ‘Script 2’) methods, measured by average number of test target coverage by each compound.

Figure S4. Promiscuity of Cutoff-40 compounds selected by fingerprint-based and bioactivity-based (using ‘Script 2’) methods, measured by average number of test target coverage by each compound.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.