Add cv csvs paper #191
Add cv csvs paper #191
Conversation
There was a problem hiding this comment.
Thanks for the PR :) I left few comments, and these are additional remarks:
- Remove the two init files added with this PR
- Clarify (ideally with files names) the difference between MLP.py, SeqBased_model.py, mlp_baseline.py
- In 6_test_cases, it's really hard to get which script does what, and deeprank2 files are still there (should be removed). I'd try to make the stucture of the folder clearer
| @@ -0,0 +1,13 @@ | |||
| #!/bin/bash | |||
There was a problem hiding this comment.
What is this file? I think it should't be here. Also, the structure in the paper branch is the following:
, please move the files accordingly
| @@ -46,12 +46,20 @@ | |||
| arg_parser.add_argument("--csv-file", "-f", | |||
| help="Name of the csv file in data/external/processed containing the cluster column. \n \ | |||
| Works as a train and validation set if provided with --test-csv.", | |||
| default="/home/daqop/mountpoint_snellius/3D-Vac/data/external/processed/all_hla_j_4.csv" | |||
| default=False, #"/home/daqop/mountpoint_snellius/3D-Vac/data/external/processed/all_hla_j_4.csv" | |||
There was a problem hiding this comment.
What is the difference between this file and MLP.py?
| import torch | ||
|
|
||
|
|
||
| def parse_pdb_dataset(folder, return_tuple=False, one_hot=False, residue_level=True, radius_pocket=10, elements=['C', 'N', 'O', 'S', 'P'], exclude_elements=['H']): |
There was a problem hiding this comment.
We have a folder called 6_test_cases, I think this file should be moved there
|
|
||
| from egnn import EGNNModel | ||
| from data_proccess_fn import data_process_fn | ||
|
|
There was a problem hiding this comment.
As above, I think this file should be move to the folder 6_test_cases
| @@ -0,0 +1,355 @@ | |||
| import numpy as np | |||
There was a problem hiding this comment.
I think that the train file should not be in this folder, that should have test-related scripts only
| @@ -0,0 +1,675 @@ | |||
| { | |||
There was a problem hiding this comment.
We don't need these two subfolders (manuscript and code). We can keep only only subfolder and call it in a more descriptive way, like paper_plots. Make sure that all the notebooks used for the actual paper's plots are there. In general, the two notebooks in code could be made clearer (even just by adding meaningful subsections title)
…ally switched with labels). Fix CNN output (previously including both test and validation)
No description provided.