question regarding encoder-decoder

[orgw]

Hi longlongman,

I've tried using your encoder-decoder on CASP-ligands but wasn't successful in fully decoding(recovering) the ligands after encoding (molecule cut off, some atoms missing)

I'm suspecting whether if it's because the patch size for the shapepretrainingencoder is too small? or maybe due to maxlen_coef being too small?

any ideas on getting the full molecule?

thanks

[longlongman]

Both two statements (small patch size & small maxlen_coef) can be the reasons. For the patch size, you can not adjust it easily because it is bound to the model parameter. The only way to enlarge the patch size is to train a new model. For maxlen_coef, you can set any value you want without re-training.

p.s. theoretically speaking, because the model is based on deep learning, it can not be 100% sure to give you the full molecule.