Multi-modal Molecule Structure-text Model for Text-based Editing and Retrieval, Nat Mach Intell 2023 (https://www.nature.com/articles/s42256-023-00759-6)
Hello, I am a senior student and I am new in the field of deep learning.
For zero-shot structure-text retrieval, I have basically reproduced the results for the SMILES modality correctly.
This is the result I get using the DrugBank Description(given structure) for the SMILES modality. This might be slight deviations, but it's essentially correct.
However, the results I get from the Graph modality on the same dataset vary significantly.
I followed the instructions in README.md and ran the code step by step. I replaced the experiment's checkpoint with the one mentioned in scripts/README.md, such as using this code:
Hello, I am a senior student and I am new in the field of deep learning. For zero-shot structure-text retrieval, I have basically reproduced the results for the SMILES modality correctly.
This is the result I get using the DrugBank Description(given structure) for the SMILES modality. This might be slight deviations, but it's essentially correct.
However, the results I get from the Graph modality on the same dataset vary significantly.
I followed the instructions in README.md and ran the code step by step. I replaced the experiment's checkpoint with the one mentioned in scripts/README.md, such as using this code:
python downstream_01_retrieval_Description_Pharmacodynamics.py \ --task=molecule_description_removed_PubChem \ --molecule_type=Graph \ --input_model_dir=../data/pretrained_MoleculeSTM/SciBERT-Graph-3e-5-1-1e-4-1-InfoNCE-0.1-32-32
I would greatly appreciate it if you could respond whenever you have time to let me know where the issue might be. Looking forward to your reply.
Also, I would like to ask what the ‘seed’ are set for all the three experiments?