ADOPT

ADOPT copied to clipboard

Retrain inference model with 1280 dimension

CFisicaro

Download of several databases when model type just esm1b

1

I ran ```bash export NEW_PROT_FASTA_FILE_PATH='/raid/app/oliver/tmp/disprot_2022_03.fasta' export NEW_PROT_RES_REPR_DIR_PATH='/raid/app/oliver/tmp/' ``` then the script (last bit of https://github.com/PeptoneInc/ADOPT/blob/main/scripts/adopt_inference.sh) ```bash TRAIN_STRATEGY="train_on_cleared_1325_test_on_117_residue_split" MODEL_TYPE="esm-1b" PRED_Z_FILE_PATH="predicted_z_scores.json" printf "Extracting residue level representation of %s \n" $NEW_PROT_FASTA_FILE_PATH python adopt/embedding.py...

oliverdutton

CFisicaro

CFisicaro

CFisicaro

CFisicaro

CFisicaro

In `inference.py` we currently have: - `torch.load(str(repr_path) + "/" + ix + ".pt")["representations"]` for `esm-msa` - `torch.load(str(repr_path) + "/" + ix + ".pt")["representations"][33]` for the other models

CFisicaro

CFisicaro

Add viz routines

1

Viz routines for paper reproducibility

CFisicaro