Hi Alejandro,

I'm a developed of CReM and I noticed that you use a fragment database with suffix sc2 in examples in documentation. I would not recommend to use it, use the one with suffix sa2. SC scores are not very reasonable to estimate synthetic accessibility and those databases were left for historical reasons. Using sa2 database will result in more meaningful molecules. Just a recommendation from my experience.
You may also ask questions by email, if you will have them - [email protected].

Pavel.

TypeError Traceback (most recent call last)
Input In [2], in <cell line: 2>()
1 from rdkit import Chem
----> 2 from moldrug import utils, fitness
3 from moldrug.data import receptors, ligands, boxes
4 import tempfile, os, gzip, shutil, requests

File ~/anaconda3/envs/py38_rdkit/lib/python3.8/site-packages/moldrug/utils.py:101, in
98 else:
99 return pdbqt_string
--> 101 def get_sim(ms:list[Chem.rdchem.Mol], ref_fps:list):
102 """Get the molecules with higher similarity to each member of ref_fps.
103
104 Parameters
(...)
114 A list of molecules with the higher similarity with their corresponded ref_fps value.
115 """
116 output = []

TypeError: 'type' object is not subscriptable