Skip to main content

Table 1 Molecular diversity assessed via the number of unique Bemis-Murcko scaffolds at the intersection between datasets

From: Molecular generation by Fast Assembly of (Deep)SMILES fragments

 

ChEMBL_train (100k)

ChEMBL_gene (100k)

TCM_train (20k)

TCM_gene (20k)

ChEMBL_train (100k)

25135

8466 (23957 new \(\sim =\) 74%)

979

982

ChEMBL_gene (100k)

8466 (23957 new \(\sim =\) 74%)

32423

802

891

TCM_train (20k)

979

802

6056

2713 (5726 new \(\sim =\) 68%)

TCM_gene (20k)

982

891

2713 (5726 new \(\sim =\) 68%)

8439