Spaces:
Running
Running
ZhaohanM
commited on
Commit
•
a1af661
1
Parent(s):
357103f
FusionGDA
Browse files- .gitattributes +3 -0
- .ipynb_checkpoints/app-checkpoint.py +64 -0
- .ipynb_checkpoints/model-checkpoint.sh +24 -0
- .ipynb_checkpoints/requirements-checkpoint.txt +7 -0
- app.py +10 -6
- save_model_ckp/gda_infoNCE_2024_GPU3090/step_300_model.bin → data/downstream/C0002395_disease.csv +2 -2
- data/downstream/GDA_Data/train.csv +3 -0
- data/downstream/GDA_Data/valid.csv +3 -0
- src/finetune/.ipynb_checkpoints/finetune-checkpoint.py +416 -0
- src/utils/__pycache__/data_loader.cpython-38.pyc +0 -0
- src/utils/__pycache__/downstream_disgenet.cpython-38.pyc +0 -0
- src/utils/__pycache__/gd_model.cpython-38.pyc +0 -0
- src/utils/__pycache__/metric_learning_models.cpython-38.pyc +0 -0
- src/utils/data_loader.py +328 -0
- src/utils/downstream_disgenet.py +3 -3
- src/utils/gd_model.py +77 -0
.gitattributes
CHANGED
@@ -35,3 +35,6 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
|
|
35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
36 |
step_300_model.bin filter=lfs diff=lfs merge=lfs -text
|
37 |
disgenet_latest.csv filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
|
|
35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
36 |
step_300_model.bin filter=lfs diff=lfs merge=lfs -text
|
37 |
disgenet_latest.csv filter=lfs diff=lfs merge=lfs -text
|
38 |
+
train.csv filter=lfs diff=lfs merge=lfs -text
|
39 |
+
valid.csv filter=lfs diff=lfs merge=lfs -text
|
40 |
+
C0002395_disease.csv filter=lfs diff=lfs merge=lfs -text
|
.ipynb_checkpoints/app-checkpoint.py
ADDED
@@ -0,0 +1,64 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
# -*- coding: utf-8 -*-
|
2 |
+
import gradio as gr
|
3 |
+
import pandas as pd
|
4 |
+
import os
|
5 |
+
import subprocess
|
6 |
+
|
7 |
+
def predict_top_100_genes(disease_id):
|
8 |
+
# Initialize paths
|
9 |
+
input_csv_path = 'data/downstream/{}_disease.csv'.format(disease_id)
|
10 |
+
output_csv_path = 'data/downstream/{}_top100.csv'.format(disease_id)
|
11 |
+
|
12 |
+
# Check if the output CSV already exists
|
13 |
+
if not os.path.exists(output_csv_path):
|
14 |
+
# Proceed with your existing code if the output file doesn't exist
|
15 |
+
df = pd.read_csv('data/pretrain/disgenet_latest.csv')
|
16 |
+
df = df[df['proteinSeq'].notna()]
|
17 |
+
|
18 |
+
# Check if the disease_id is present in the dataframe
|
19 |
+
if disease_id not in df['diseaseId'].values:
|
20 |
+
return f"Error: Disease ID '{disease_id}' not found in the database. Please check the ID and try again."
|
21 |
+
|
22 |
+
desired_diseaseDes = df[df['diseaseId'] == disease_id]['diseaseDes'].iloc[0]
|
23 |
+
related_proteins = df[df['diseaseDes'] == desired_diseaseDes]['proteinSeq'].unique()
|
24 |
+
df['score'] = df['proteinSeq'].isin(related_proteins).astype(int)
|
25 |
+
new_df = pd.DataFrame({
|
26 |
+
'diseaseId': disease_id,
|
27 |
+
'diseaseDes': desired_diseaseDes,
|
28 |
+
'geneSymbol': df['geneSymbol'],
|
29 |
+
'proteinSeq': df['proteinSeq'],
|
30 |
+
'score': df['score']
|
31 |
+
}).drop_duplicates().reset_index(drop=True)
|
32 |
+
|
33 |
+
new_df.to_csv(input_csv_path, index=False)
|
34 |
+
|
35 |
+
# Call the model script only if the output CSV does not exist
|
36 |
+
script_path = 'model.sh'
|
37 |
+
subprocess.run(['bash', script_path, input_csv_path, output_csv_path], check=True)
|
38 |
+
|
39 |
+
# Read the model output file or the existing file to get the top 100 genes
|
40 |
+
output_df = pd.read_csv(output_csv_path)
|
41 |
+
# Update here to select only the required columns and rename them
|
42 |
+
result_df = output_df[['geneSymbol', 'Prediction_score']].rename(columns={'geneSymbol': 'Gene', 'Prediction_score': 'Score'}).head(100)
|
43 |
+
|
44 |
+
return result_df
|
45 |
+
|
46 |
+
iface = gr.Interface(
|
47 |
+
fn=predict_top_100_genes,
|
48 |
+
inputs=gr.Textbox(lines=1, placeholder="Enter Disease ID Here...", label="Disease ID"),
|
49 |
+
outputs=gr.Dataframe(label="Predicted Top 100 Related Genes"),
|
50 |
+
title="Gene Disease Association Prediction",
|
51 |
+
description = (
|
52 |
+
"This AI model predicts the top 100 genes associated with a given disease based on 16,733 genes."
|
53 |
+
" To get started, you need a Disease ID (UMLS CUI), which can be obtained from the DisGeNET database. "
|
54 |
+
"\n\n**Steps to Obtain a Disease ID from DisGeNET:**\n"
|
55 |
+
"1. Visit the DisGeNET website: [https://www.disgenet.org/search](https://www.disgenet.org/search).\n"
|
56 |
+
"2. Use the search bar to enter your disease of interest. For instance, if you're interested in 'Alzheimer's Disease', type 'Alzheimer's Disease' into the search bar.\n"
|
57 |
+
"3. From the search results, identify the disease you're researching. The Disease ID (UMLS CUI) is listed alongside each disease name, e.g. C0002395.\n"
|
58 |
+
"4. Enter the Disease ID into the input box below and submit.\n\n"
|
59 |
+
"The DisGeNET database contains all known gene-disease associations and associated evidence. In addition, it is able to find the corresponding diseases based on a gene.\n"
|
60 |
+
"\n**The model will take about 18 minutes to inference a new disease.**\n"
|
61 |
+
)
|
62 |
+
)
|
63 |
+
|
64 |
+
iface.launch(share=True)
|
.ipynb_checkpoints/model-checkpoint.sh
ADDED
@@ -0,0 +1,24 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
#!/bin/bash
|
2 |
+
|
3 |
+
input_csv_path="$1"
|
4 |
+
output_csv_path="$2"
|
5 |
+
max_depth=6
|
6 |
+
device='cuda:0'
|
7 |
+
model_path_list=(
|
8 |
+
"../../save_model_ckp/gda_infoNCE_2024_GPU3090" \
|
9 |
+
)
|
10 |
+
|
11 |
+
cd ../src/finetune/
|
12 |
+
for save_model_path in ${model_path_list[@]}; do
|
13 |
+
num_leaves=$((2**($max_depth-1)))
|
14 |
+
python finetune.py \
|
15 |
+
--input_csv_path $input_csv_path \
|
16 |
+
--output_csv_path $output_csv_path \
|
17 |
+
--save_model_path $save_model_path \
|
18 |
+
--device $device \
|
19 |
+
--batch_size 128 \
|
20 |
+
--step "300" \
|
21 |
+
--use_pooled \
|
22 |
+
--num_leaves $num_leaves \
|
23 |
+
--max_depth $max_depth
|
24 |
+
done
|
.ipynb_checkpoints/requirements-checkpoint.txt
ADDED
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
lightgbm
|
2 |
+
pytorch-metric-learning
|
3 |
+
torch
|
4 |
+
transformers
|
5 |
+
PyTDC
|
6 |
+
gradio
|
7 |
+
numpy
|
app.py
CHANGED
@@ -6,14 +6,19 @@ import subprocess
|
|
6 |
|
7 |
def predict_top_100_genes(disease_id):
|
8 |
# Initialize paths
|
9 |
-
input_csv_path = '
|
10 |
-
output_csv_path = '
|
11 |
|
12 |
# Check if the output CSV already exists
|
13 |
if not os.path.exists(output_csv_path):
|
14 |
# Proceed with your existing code if the output file doesn't exist
|
15 |
-
df = pd.read_csv('
|
16 |
df = df[df['proteinSeq'].notna()]
|
|
|
|
|
|
|
|
|
|
|
17 |
desired_diseaseDes = df[df['diseaseId'] == disease_id]['diseaseDes'].iloc[0]
|
18 |
related_proteins = df[df['diseaseDes'] == desired_diseaseDes]['proteinSeq'].unique()
|
19 |
df['score'] = df['proteinSeq'].isin(related_proteins).astype(int)
|
@@ -38,7 +43,6 @@ def predict_top_100_genes(disease_id):
|
|
38 |
|
39 |
return result_df
|
40 |
|
41 |
-
|
42 |
iface = gr.Interface(
|
43 |
fn=predict_top_100_genes,
|
44 |
inputs=gr.Textbox(lines=1, placeholder="Enter Disease ID Here...", label="Disease ID"),
|
@@ -54,7 +58,7 @@ iface = gr.Interface(
|
|
54 |
"4. Enter the Disease ID into the input box below and submit.\n\n"
|
55 |
"The DisGeNET database contains all known gene-disease associations and associated evidence. In addition, it is able to find the corresponding diseases based on a gene.\n"
|
56 |
"\n**The model will take about 18 minutes to inference a new disease.**\n"
|
57 |
-
)
|
58 |
)
|
59 |
|
60 |
-
iface.launch(share=True)
|
|
|
6 |
|
7 |
def predict_top_100_genes(disease_id):
|
8 |
# Initialize paths
|
9 |
+
input_csv_path = 'data/downstream/{}_disease.csv'.format(disease_id)
|
10 |
+
output_csv_path = 'data/downstream/{}_top100.csv'.format(disease_id)
|
11 |
|
12 |
# Check if the output CSV already exists
|
13 |
if not os.path.exists(output_csv_path):
|
14 |
# Proceed with your existing code if the output file doesn't exist
|
15 |
+
df = pd.read_csv('data/pretrain/disgenet_latest.csv')
|
16 |
df = df[df['proteinSeq'].notna()]
|
17 |
+
|
18 |
+
# Check if the disease_id is present in the dataframe
|
19 |
+
if disease_id not in df['diseaseId'].values:
|
20 |
+
return f"Error: Disease ID '{disease_id}' not found in the database. Please check the ID and try again."
|
21 |
+
|
22 |
desired_diseaseDes = df[df['diseaseId'] == disease_id]['diseaseDes'].iloc[0]
|
23 |
related_proteins = df[df['diseaseDes'] == desired_diseaseDes]['proteinSeq'].unique()
|
24 |
df['score'] = df['proteinSeq'].isin(related_proteins).astype(int)
|
|
|
43 |
|
44 |
return result_df
|
45 |
|
|
|
46 |
iface = gr.Interface(
|
47 |
fn=predict_top_100_genes,
|
48 |
inputs=gr.Textbox(lines=1, placeholder="Enter Disease ID Here...", label="Disease ID"),
|
|
|
58 |
"4. Enter the Disease ID into the input box below and submit.\n\n"
|
59 |
"The DisGeNET database contains all known gene-disease associations and associated evidence. In addition, it is able to find the corresponding diseases based on a gene.\n"
|
60 |
"\n**The model will take about 18 minutes to inference a new disease.**\n"
|
61 |
+
)
|
62 |
)
|
63 |
|
64 |
+
iface.launch(share=True)
|
save_model_ckp/gda_infoNCE_2024_GPU3090/step_300_model.bin → data/downstream/C0002395_disease.csv
RENAMED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
-
size
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:2f7b5562dceae680af5fbe305e06e5ebacafb9bf8404ecebc04b8ecc60a3495d
|
3 |
+
size 44085860
|
data/downstream/GDA_Data/train.csv
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:d0533480255f46747ca973110aaa031892bdfd5ca9b2f9bc989f91ce893385a2
|
3 |
+
size 117023981
|
data/downstream/GDA_Data/valid.csv
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:03c90fd29ac0af5370c8dc66317cab3de004c1771f47f91301fcb6c11204815f
|
3 |
+
size 29321915
|
src/finetune/.ipynb_checkpoints/finetune-checkpoint.py
ADDED
@@ -0,0 +1,416 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import argparse
|
2 |
+
import os
|
3 |
+
import random
|
4 |
+
import string
|
5 |
+
import sys
|
6 |
+
import pandas as pd
|
7 |
+
from datetime import datetime
|
8 |
+
|
9 |
+
sys.path.append("../")
|
10 |
+
import numpy as np
|
11 |
+
import torch
|
12 |
+
import lightgbm as lgb
|
13 |
+
import sklearn.metrics as metrics
|
14 |
+
from sklearn.utils import class_weight
|
15 |
+
from sklearn.model_selection import train_test_split
|
16 |
+
from sklearn.metrics import accuracy_score, precision_recall_curve, f1_score, precision_recall_fscore_support,roc_auc_score
|
17 |
+
from torch.utils.data import DataLoader
|
18 |
+
from tqdm.auto import tqdm
|
19 |
+
from transformers import EsmTokenizer, EsmForMaskedLM, BertModel, BertTokenizer, AutoTokenizer, EsmModel
|
20 |
+
from utils.downstream_disgenet import DisGeNETProcessor
|
21 |
+
from utils.metric_learning_models import GDA_Metric_Learning
|
22 |
+
|
23 |
+
def parse_config():
|
24 |
+
parser = argparse.ArgumentParser()
|
25 |
+
parser.add_argument('-f')
|
26 |
+
parser.add_argument("--step", type=int, default=0)
|
27 |
+
parser.add_argument(
|
28 |
+
"--save_model_path",
|
29 |
+
type=str,
|
30 |
+
default=None,
|
31 |
+
help="path of the pretrained disease model located",
|
32 |
+
)
|
33 |
+
parser.add_argument(
|
34 |
+
"--prot_encoder_path",
|
35 |
+
type=str,
|
36 |
+
default="facebook/esm2_t33_650M_UR50D",
|
37 |
+
#"facebook/galactica-6.7b", "Rostlab/prot_bert" “facebook/esm2_t33_650M_UR50D”
|
38 |
+
help="path/name of protein encoder model located",
|
39 |
+
)
|
40 |
+
parser.add_argument(
|
41 |
+
"--disease_encoder_path",
|
42 |
+
type=str,
|
43 |
+
default="microsoft/BiomedNLP-PubMedBERT-base-uncased-abstract-fulltext",
|
44 |
+
help="path/name of textual pre-trained language model",
|
45 |
+
)
|
46 |
+
parser.add_argument("--reduction_factor", type=int, default=8)
|
47 |
+
parser.add_argument(
|
48 |
+
"--loss",
|
49 |
+
help="{ms_loss|infoNCE|cosine_loss|circle_loss|triplet_loss}}",
|
50 |
+
default="infoNCE",
|
51 |
+
)
|
52 |
+
parser.add_argument(
|
53 |
+
"--input_feature_save_path",
|
54 |
+
type=str,
|
55 |
+
default="../../data/processed_disease",
|
56 |
+
help="path of tokenized training data",
|
57 |
+
)
|
58 |
+
parser.add_argument(
|
59 |
+
"--agg_mode", default="mean_all_tok", type=str, help="{cls|mean|mean_all_tok}"
|
60 |
+
)
|
61 |
+
parser.add_argument("--batch_size", type=int, default=256)
|
62 |
+
parser.add_argument("--patience", type=int, default=5)
|
63 |
+
parser.add_argument("--num_leaves", type=int, default=5)
|
64 |
+
parser.add_argument("--max_depth", type=int, default=5)
|
65 |
+
parser.add_argument("--lr", type=float, default=0.35)
|
66 |
+
parser.add_argument("--dropout", type=float, default=0.1)
|
67 |
+
parser.add_argument("--test", type=int, default=0)
|
68 |
+
parser.add_argument("--use_miner", action="store_true")
|
69 |
+
parser.add_argument("--miner_margin", default=0.2, type=float)
|
70 |
+
parser.add_argument("--freeze_prot_encoder", action="store_true")
|
71 |
+
parser.add_argument("--freeze_disease_encoder", action="store_true")
|
72 |
+
parser.add_argument("--use_adapter", action="store_true")
|
73 |
+
parser.add_argument("--use_pooled", action="store_true")
|
74 |
+
parser.add_argument("--device", type=str, default="cpu")
|
75 |
+
parser.add_argument(
|
76 |
+
"--use_both_feature",
|
77 |
+
help="use the both features of gnn_feature_v1_samples and pretrained models",
|
78 |
+
action="store_true",
|
79 |
+
)
|
80 |
+
parser.add_argument(
|
81 |
+
"--use_v1_feature_only",
|
82 |
+
help="use the features of gnn_feature_v1_samples only",
|
83 |
+
action="store_true",
|
84 |
+
)
|
85 |
+
parser.add_argument(
|
86 |
+
"--save_path_prefix",
|
87 |
+
type=str,
|
88 |
+
default="../../save_model_ckp/finetune/",
|
89 |
+
help="save the result in which directory",
|
90 |
+
)
|
91 |
+
parser.add_argument(
|
92 |
+
"--save_name", default="fine_tune", type=str, help="the name of the saved file"
|
93 |
+
)
|
94 |
+
# Add argument for input CSV file path
|
95 |
+
parser.add_argument("--input_csv_path", type=str, required=True, help="Path to the input CSV file.")
|
96 |
+
|
97 |
+
# Add argument for output CSV file path
|
98 |
+
parser.add_argument("--output_csv_path", type=str, required=True, help="Path to the output CSV file.")
|
99 |
+
return parser.parse_args()
|
100 |
+
|
101 |
+
def get_feature(model, dataloader, args):
|
102 |
+
x = list()
|
103 |
+
y = list()
|
104 |
+
with torch.no_grad():
|
105 |
+
for step, batch in tqdm(enumerate(dataloader)):
|
106 |
+
prot_input_ids, prot_attention_mask, dis_input_ids, dis_attention_mask, y1 = batch
|
107 |
+
prot_input = {
|
108 |
+
'input_ids': prot_input_ids.to(args.device),
|
109 |
+
'attention_mask': prot_attention_mask.to(args.device)
|
110 |
+
}
|
111 |
+
dis_input = {
|
112 |
+
'input_ids': dis_input_ids.to(args.device),
|
113 |
+
'attention_mask': dis_attention_mask.to(args.device)
|
114 |
+
}
|
115 |
+
feature_output = model.predict(prot_input, dis_input)
|
116 |
+
x1 = feature_output.cpu().numpy()
|
117 |
+
x.append(x1)
|
118 |
+
y.append(y1.cpu().numpy())
|
119 |
+
x = np.concatenate(x, axis=0)
|
120 |
+
y = np.concatenate(y, axis=0)
|
121 |
+
return x, y
|
122 |
+
|
123 |
+
|
124 |
+
def encode_pretrained_feature(args, disGeNET):
|
125 |
+
input_feat_file = os.path.join(
|
126 |
+
args.input_feature_save_path,
|
127 |
+
f"{args.model_short}_{args.step}_use_{'pooled' if args.use_pooled else 'cls'}_feat.npz",
|
128 |
+
)
|
129 |
+
|
130 |
+
if os.path.exists(input_feat_file):
|
131 |
+
print(f"load prior feature data from {input_feat_file}.")
|
132 |
+
loaded = np.load(input_feat_file)
|
133 |
+
x_train, y_train = loaded["x_train"], loaded["y_train"]
|
134 |
+
x_valid, y_valid = loaded["x_valid"], loaded["y_valid"]
|
135 |
+
# x_test, y_test = loaded["x_test"], loaded["y_test"]
|
136 |
+
|
137 |
+
prot_tokenizer = EsmTokenizer.from_pretrained(args.prot_encoder_path, do_lower_case=False)
|
138 |
+
# prot_tokenizer = BertTokenizer.from_pretrained(args.prot_encoder_path, do_lower_case=False)
|
139 |
+
print("prot_tokenizer", len(prot_tokenizer))
|
140 |
+
disease_tokenizer = BertTokenizer.from_pretrained(args.disease_encoder_path)
|
141 |
+
print("disease_tokenizer", len(disease_tokenizer))
|
142 |
+
|
143 |
+
prot_model = EsmModel.from_pretrained(args.prot_encoder_path)
|
144 |
+
# prot_model = BertModel.from_pretrained(args.prot_encoder_path)
|
145 |
+
disease_model = BertModel.from_pretrained(args.disease_encoder_path)
|
146 |
+
|
147 |
+
if args.save_model_path:
|
148 |
+
model = GDA_Metric_Learning(prot_model, disease_model, 1280, 768, args)
|
149 |
+
|
150 |
+
if args.use_adapter:
|
151 |
+
prot_model_path = os.path.join(
|
152 |
+
args.save_model_path, f"prot_adapter_step_{args.step}"
|
153 |
+
)
|
154 |
+
disease_model_path = os.path.join(
|
155 |
+
args.save_model_path, f"disease_adapter_step_{args.step}"
|
156 |
+
)
|
157 |
+
model.load_adapters(prot_model_path, disease_model_path)
|
158 |
+
else:
|
159 |
+
prot_model_path = os.path.join(
|
160 |
+
args.save_model_path, f"step_{args.step}_model.bin"
|
161 |
+
)# , f"step_{args.step}_model.bin"
|
162 |
+
disease_model_path = os.path.join(
|
163 |
+
args.save_model_path, f"step_{args.step}_model.bin"
|
164 |
+
)
|
165 |
+
model.non_adapters(prot_model_path, disease_model_path)
|
166 |
+
|
167 |
+
model = model.to(args.device)
|
168 |
+
prot_model = model.prot_encoder
|
169 |
+
disease_model = model.disease_encoder
|
170 |
+
print(f"loaded prior model {args.save_model_path}.")
|
171 |
+
|
172 |
+
def collate_fn_batch_encoding(batch):
|
173 |
+
query1, query2, scores = zip(*batch)
|
174 |
+
|
175 |
+
query_encodings1 = prot_tokenizer.batch_encode_plus(
|
176 |
+
list(query1),
|
177 |
+
max_length=512,
|
178 |
+
padding="max_length",
|
179 |
+
truncation=True,
|
180 |
+
add_special_tokens=True,
|
181 |
+
return_tensors="pt",
|
182 |
+
)
|
183 |
+
query_encodings2 = disease_tokenizer.batch_encode_plus(
|
184 |
+
list(query2),
|
185 |
+
max_length=512,
|
186 |
+
padding="max_length",
|
187 |
+
truncation=True,
|
188 |
+
add_special_tokens=True,
|
189 |
+
return_tensors="pt",
|
190 |
+
)
|
191 |
+
scores = torch.tensor(list(scores))
|
192 |
+
attention_mask1 = query_encodings1["attention_mask"].bool()
|
193 |
+
attention_mask2 = query_encodings2["attention_mask"].bool()
|
194 |
+
|
195 |
+
return query_encodings1["input_ids"], attention_mask1, query_encodings2["input_ids"], attention_mask2, scores
|
196 |
+
|
197 |
+
test_examples = disGeNET.get_test_examples(args.test)
|
198 |
+
print(f"get test examples: {len(test_examples)}")
|
199 |
+
|
200 |
+
test_dataloader = DataLoader(
|
201 |
+
test_examples,
|
202 |
+
batch_size=args.batch_size,
|
203 |
+
shuffle=False,
|
204 |
+
collate_fn=collate_fn_batch_encoding,
|
205 |
+
)
|
206 |
+
print( f"dataset loaded: test-{len(test_examples)}")
|
207 |
+
|
208 |
+
x_test, y_test = get_feature(model, test_dataloader, args)
|
209 |
+
|
210 |
+
else:
|
211 |
+
prot_tokenizer = EsmTokenizer.from_pretrained(args.prot_encoder_path, do_lower_case=False)
|
212 |
+
# prot_tokenizer = BertTokenizer.from_pretrained(args.prot_encoder_path, do_lower_case=False)
|
213 |
+
print("prot_tokenizer", len(prot_tokenizer))
|
214 |
+
disease_tokenizer = BertTokenizer.from_pretrained(args.disease_encoder_path)
|
215 |
+
print("disease_tokenizer", len(disease_tokenizer))
|
216 |
+
|
217 |
+
prot_model = EsmModel.from_pretrained(args.prot_encoder_path)
|
218 |
+
# prot_model = BertModel.from_pretrained(args.prot_encoder_path)
|
219 |
+
disease_model = BertModel.from_pretrained(args.disease_encoder_path)
|
220 |
+
|
221 |
+
if args.save_model_path:
|
222 |
+
model = GDA_Metric_Learning(prot_model, disease_model, 1280, 768, args)
|
223 |
+
|
224 |
+
if args.use_adapter:
|
225 |
+
prot_model_path = os.path.join(
|
226 |
+
args.save_model_path, f"prot_adapter_step_{args.step}"
|
227 |
+
)
|
228 |
+
disease_model_path = os.path.join(
|
229 |
+
args.save_model_path, f"disease_adapter_step_{args.step}"
|
230 |
+
)
|
231 |
+
model.load_adapters(prot_model_path, disease_model_path)
|
232 |
+
else:
|
233 |
+
prot_model_path = os.path.join(
|
234 |
+
args.save_model_path, f"step_{args.step}_model.bin"
|
235 |
+
)# , f"step_{args.step}_model.bin"
|
236 |
+
disease_model_path = os.path.join(
|
237 |
+
args.save_model_path, f"step_{args.step}_model.bin"
|
238 |
+
)
|
239 |
+
model.non_adapters(prot_model_path, disease_model_path)
|
240 |
+
|
241 |
+
model = model.to(args.device)
|
242 |
+
prot_model = model.prot_encoder
|
243 |
+
disease_model = model.disease_encoder
|
244 |
+
print(f"loaded prior model {args.save_model_path}.")
|
245 |
+
|
246 |
+
def collate_fn_batch_encoding(batch):
|
247 |
+
query1, query2, scores = zip(*batch)
|
248 |
+
|
249 |
+
query_encodings1 = prot_tokenizer.batch_encode_plus(
|
250 |
+
list(query1),
|
251 |
+
max_length=512,
|
252 |
+
padding="max_length",
|
253 |
+
truncation=True,
|
254 |
+
add_special_tokens=True,
|
255 |
+
return_tensors="pt",
|
256 |
+
)
|
257 |
+
query_encodings2 = disease_tokenizer.batch_encode_plus(
|
258 |
+
list(query2),
|
259 |
+
max_length=512,
|
260 |
+
padding="max_length",
|
261 |
+
truncation=True,
|
262 |
+
add_special_tokens=True,
|
263 |
+
return_tensors="pt",
|
264 |
+
)
|
265 |
+
scores = torch.tensor(list(scores))
|
266 |
+
attention_mask1 = query_encodings1["attention_mask"].bool()
|
267 |
+
attention_mask2 = query_encodings2["attention_mask"].bool()
|
268 |
+
|
269 |
+
return query_encodings1["input_ids"], attention_mask1, query_encodings2["input_ids"], attention_mask2, scores
|
270 |
+
|
271 |
+
train_examples = disGeNET.get_train_examples(args.test)
|
272 |
+
print(f"get training examples: {len(train_examples)}")
|
273 |
+
valid_examples = disGeNET.get_val_examples(args.test)
|
274 |
+
print(f"get validation examples: {len(valid_examples)}")
|
275 |
+
test_examples = disGeNET.get_test_examples(args.test)
|
276 |
+
print(f"get test examples: {len(test_examples)}")
|
277 |
+
|
278 |
+
train_dataloader = DataLoader(
|
279 |
+
train_examples,
|
280 |
+
batch_size=args.batch_size,
|
281 |
+
shuffle=False,
|
282 |
+
collate_fn=collate_fn_batch_encoding,
|
283 |
+
)
|
284 |
+
valid_dataloader = DataLoader(
|
285 |
+
valid_examples,
|
286 |
+
batch_size=args.batch_size,
|
287 |
+
shuffle=False,
|
288 |
+
collate_fn=collate_fn_batch_encoding,
|
289 |
+
)
|
290 |
+
test_dataloader = DataLoader(
|
291 |
+
test_examples,
|
292 |
+
batch_size=args.batch_size,
|
293 |
+
shuffle=False,
|
294 |
+
collate_fn=collate_fn_batch_encoding,
|
295 |
+
)
|
296 |
+
print( f"dataset loaded: train-{len(train_examples)}; valid-{len(valid_examples)}; test-{len(test_examples)}")
|
297 |
+
|
298 |
+
x_train, y_train = get_feature(model, train_dataloader, args)
|
299 |
+
x_valid, y_valid = get_feature(model, valid_dataloader, args)
|
300 |
+
x_test, y_test = get_feature(model, test_dataloader, args)
|
301 |
+
|
302 |
+
# Save input feature to reduce encoding time
|
303 |
+
np.savez_compressed(
|
304 |
+
input_feat_file,
|
305 |
+
x_train=x_train,
|
306 |
+
y_train=y_train,
|
307 |
+
x_valid=x_valid,
|
308 |
+
y_valid=y_valid,
|
309 |
+
)
|
310 |
+
print(f"save input feature into {input_feat_file}")
|
311 |
+
# Save input feature to reduce encoding time
|
312 |
+
return x_train, y_train, x_valid, y_valid, x_test, y_test
|
313 |
+
|
314 |
+
|
315 |
+
def train(args):
|
316 |
+
# defining parameters
|
317 |
+
if args.save_model_path:
|
318 |
+
args.model_short = (
|
319 |
+
args.save_model_path.split("/")[-1]
|
320 |
+
)
|
321 |
+
print(f"model name {args.model_short}")
|
322 |
+
|
323 |
+
else:
|
324 |
+
args.model_short = (
|
325 |
+
args.disease_encoder_path.split("/")[-1]
|
326 |
+
)
|
327 |
+
print(f"model name {args.model_short}")
|
328 |
+
|
329 |
+
# disGeNET = DisGeNETProcessor()
|
330 |
+
disGeNET = DisGeNETProcessor(input_csv_path=args.input_csv_path)
|
331 |
+
|
332 |
+
|
333 |
+
x_train, y_train, x_valid, y_valid, x_test, y_test = encode_pretrained_feature(args, disGeNET)
|
334 |
+
|
335 |
+
print("train: ", x_train.shape, y_train.shape)
|
336 |
+
print("valid: ", x_valid.shape, y_valid.shape)
|
337 |
+
print("test: ", x_test.shape, y_test.shape)
|
338 |
+
|
339 |
+
params = {
|
340 |
+
"task": "train", # "predict" train
|
341 |
+
"boosting": "gbdt", # "The options are "gbdt" (traditional Gradient Boosting Decision Tree), "rf" (Random Forest), "dart" (Dropouts meet Multiple Additive Regression Trees), or "goss" (Gradient-based One-Side Sampling). The default is "gbdt"."
|
342 |
+
"objective": "binary",
|
343 |
+
"num_leaves": args.num_leaves,
|
344 |
+
"early_stopping_round": 30,
|
345 |
+
"max_depth": args.max_depth,
|
346 |
+
"learning_rate": args.lr,
|
347 |
+
"metric": "binary_logloss", #"metric": "l2","binary_logloss" "auc"
|
348 |
+
"verbose": 1,
|
349 |
+
}
|
350 |
+
|
351 |
+
lgb_train = lgb.Dataset(x_train, y_train)
|
352 |
+
lgb_valid = lgb.Dataset(x_valid, y_valid)
|
353 |
+
lgb_eval = lgb.Dataset(x_test, y_test, reference=lgb_train)
|
354 |
+
|
355 |
+
# fitting the model
|
356 |
+
model = lgb.train(
|
357 |
+
params, train_set=lgb_train, valid_sets=lgb_valid)
|
358 |
+
|
359 |
+
# prediction
|
360 |
+
valid_y_pred = model.predict(x_valid)
|
361 |
+
test_y_pred = model.predict(x_test)
|
362 |
+
|
363 |
+
# predict liver fibrosis
|
364 |
+
predictions_df = pd.DataFrame(test_y_pred, columns=["Prediction_score"])
|
365 |
+
# data_test = pd.read_csv('/nfs/dpa_pretrain/data/downstream/GDA_Data/test_tdc.csv')
|
366 |
+
data_test = pd.read_csv(args.input_csv_path)
|
367 |
+
predictions = pd.concat([data_test, predictions_df], axis=1)
|
368 |
+
# filtered_dataset = test_dataset_with_predictions[test_dataset_with_predictions['diseaseId'] == 'C0009714']
|
369 |
+
predictions.sort_values(by='Prediction_score', ascending=False, inplace=True)
|
370 |
+
top_100_predictions = predictions.head(100)
|
371 |
+
top_100_predictions.to_csv(args.output_csv_path, index=False)
|
372 |
+
|
373 |
+
# Accuracy
|
374 |
+
y_pred = model.predict(x_test, num_iteration=model.best_iteration)
|
375 |
+
y_pred[y_pred >= 0.5] = 1
|
376 |
+
y_pred[y_pred < 0.5] = 0
|
377 |
+
accuracy = accuracy_score(y_test, y_pred)
|
378 |
+
|
379 |
+
# AUC
|
380 |
+
valid_roc_auc_score = metrics.roc_auc_score(y_valid, valid_y_pred)
|
381 |
+
valid_average_precision_score = metrics.average_precision_score(
|
382 |
+
y_valid, valid_y_pred
|
383 |
+
)
|
384 |
+
test_roc_auc_score = metrics.roc_auc_score(y_test, test_y_pred)
|
385 |
+
test_average_precision_score = metrics.average_precision_score(y_test, test_y_pred)
|
386 |
+
|
387 |
+
# AUPR
|
388 |
+
valid_aupr = metrics.average_precision_score(y_valid, valid_y_pred)
|
389 |
+
test_aupr = metrics.average_precision_score(y_test, test_y_pred)
|
390 |
+
|
391 |
+
# Fmax
|
392 |
+
valid_precision, valid_recall, valid_thresholds = precision_recall_curve(y_valid, valid_y_pred)
|
393 |
+
valid_fmax = (2 * valid_precision * valid_recall / (valid_precision + valid_recall)).max()
|
394 |
+
test_precision, test_recall, test_thresholds = precision_recall_curve(y_test, test_y_pred)
|
395 |
+
test_fmax = (2 * test_precision * test_recall / (test_precision + test_recall)).max()
|
396 |
+
|
397 |
+
# F1
|
398 |
+
valid_f1 = f1_score(y_valid, valid_y_pred >= 0.5)
|
399 |
+
test_f1 = f1_score(y_test, test_y_pred >= 0.5)
|
400 |
+
|
401 |
+
|
402 |
+
if __name__ == "__main__":
|
403 |
+
args = parse_config()
|
404 |
+
if torch.cuda.is_available():
|
405 |
+
print("cuda is available.")
|
406 |
+
print(f"current device {args}.")
|
407 |
+
else:
|
408 |
+
args.device = "cpu"
|
409 |
+
timestamp_str = datetime.now().strftime("%Y%m%d_%H%M%S")
|
410 |
+
random_str = "".join([random.choice(string.ascii_lowercase) for n in range(6)])
|
411 |
+
best_model_dir = (
|
412 |
+
f"{args.save_path_prefix}{args.save_name}_{timestamp_str}_{random_str}/"
|
413 |
+
)
|
414 |
+
os.makedirs(best_model_dir)
|
415 |
+
args.save_name = best_model_dir
|
416 |
+
train(args)
|
src/utils/__pycache__/data_loader.cpython-38.pyc
ADDED
Binary file (7.09 kB). View file
|
|
src/utils/__pycache__/downstream_disgenet.cpython-38.pyc
ADDED
Binary file (2.97 kB). View file
|
|
src/utils/__pycache__/gd_model.cpython-38.pyc
ADDED
Binary file (2.84 kB). View file
|
|
src/utils/__pycache__/metric_learning_models.cpython-38.pyc
ADDED
Binary file (17.3 kB). View file
|
|
src/utils/data_loader.py
ADDED
@@ -0,0 +1,328 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import logging
|
2 |
+
import sys
|
3 |
+
import numpy as np
|
4 |
+
sys.path.append("../")
|
5 |
+
# from tdc.multi_pred import GDA
|
6 |
+
import pandas as pd
|
7 |
+
from torch.utils.data import Dataset
|
8 |
+
|
9 |
+
LOGGER = logging.getLogger(__name__)
|
10 |
+
|
11 |
+
class GDA_Dataset(Dataset):
|
12 |
+
"""
|
13 |
+
Candidate Dataset for:
|
14 |
+
ALL gene-to-disease interactions
|
15 |
+
"""
|
16 |
+
def __init__(self, data_examples):
|
17 |
+
self.protein_seqs = data_examples[0]
|
18 |
+
self.disease_dess = data_examples[1]
|
19 |
+
self.scores = data_examples[2]
|
20 |
+
|
21 |
+
def __getitem__(self, query_idx):
|
22 |
+
|
23 |
+
protein_seq = self.protein_seqs[query_idx]
|
24 |
+
disease_des = self.disease_dess[query_idx]
|
25 |
+
score = self.scores[query_idx]
|
26 |
+
|
27 |
+
return protein_seq, disease_des, score
|
28 |
+
|
29 |
+
def __len__(self):
|
30 |
+
return len(self.protein_seqs)
|
31 |
+
|
32 |
+
|
33 |
+
class TDC_Pretrain_Dataset(Dataset):
|
34 |
+
"""
|
35 |
+
Dataset of TDC:
|
36 |
+
ALL gene-disease associations
|
37 |
+
"""
|
38 |
+
def __init__(self, data_dir="../../data/pretrain/", test=False):
|
39 |
+
LOGGER.info("Initializing TDC Pretraining Dataset ! ...")
|
40 |
+
|
41 |
+
data = GDA(name="DisGeNET") # , path=data_dir
|
42 |
+
data.neg_sample(frac = 1)
|
43 |
+
data.binarize(threshold = 0, order = 'ascending')
|
44 |
+
self.datasets = data.get_split()
|
45 |
+
self.name = "DisGeNET"
|
46 |
+
self.dataset_df = self.datasets['train']
|
47 |
+
# self.dataset_df = pd.read_csv(f"{data_dir}/disgenet_gda.csv")
|
48 |
+
self.dataset_df = self.dataset_df[
|
49 |
+
["Gene", "Disease", "Y"]
|
50 |
+
].dropna() # Drop missing values.
|
51 |
+
# print(self.dataset_df.head())
|
52 |
+
print(
|
53 |
+
f"{data_dir}TDC training dataset loaded, found associations: {len(self.dataset_df.index)}"
|
54 |
+
)
|
55 |
+
self.protein_seqs = self.dataset_df["Gene"].values
|
56 |
+
self.disease_dess = self.dataset_df["Disease"].values
|
57 |
+
self.scores = len(self.dataset_df["Y"].values) * [1]
|
58 |
+
|
59 |
+
def __getitem__(self, query_idx):
|
60 |
+
|
61 |
+
protein_seq = self.protein_seqs[query_idx]
|
62 |
+
disease_des = self.disease_dess[query_idx]
|
63 |
+
score = self.scores[query_idx]
|
64 |
+
|
65 |
+
return protein_seq, disease_des, score
|
66 |
+
|
67 |
+
def __len__(self):
|
68 |
+
return len(self.protein_seqs)
|
69 |
+
|
70 |
+
class GDA_Pretrain_Dataset(Dataset):
|
71 |
+
"""
|
72 |
+
Candidate Dataset for:
|
73 |
+
ALL gene-disease associations
|
74 |
+
"""
|
75 |
+
|
76 |
+
def __init__(self, data_dir="../../data/pretrain/", test=False, split="train", val_ratio=0.2):
|
77 |
+
LOGGER.info("Initializing GDA Pretraining Dataset ! ...")
|
78 |
+
self.dataset_df = pd.read_csv(f"{data_dir}/disgenet_gda.csv")
|
79 |
+
self.dataset_df = self.dataset_df[["proteinSeq", "diseaseDes", "score"]].dropna()
|
80 |
+
self.dataset_df = self.dataset_df.sample(frac=1, random_state=42).reset_index(drop=True)
|
81 |
+
|
82 |
+
num_val_samples = int(len(self.dataset_df) * val_ratio)
|
83 |
+
if split == "train":
|
84 |
+
self.dataset_df = self.dataset_df[:-num_val_samples]
|
85 |
+
print(f"{data_dir}disgenet_gda.csv loaded, found train associations: {len(self.dataset_df.index)}")
|
86 |
+
elif split == "val":
|
87 |
+
self.dataset_df = self.dataset_df[-num_val_samples:]
|
88 |
+
print(f"{data_dir}disgenet_gda.csv loaded, found valid associations: {len(self.dataset_df.index)}")
|
89 |
+
|
90 |
+
if test:
|
91 |
+
self.protein_seqs = self.dataset_df["proteinSeq"].values[:128]
|
92 |
+
self.disease_dess = self.dataset_df["diseaseDes"].values[:128]
|
93 |
+
self.scores = 128 * [1]
|
94 |
+
else:
|
95 |
+
self.protein_seqs = self.dataset_df["proteinSeq"].values
|
96 |
+
self.disease_dess = self.dataset_df["diseaseDes"].values
|
97 |
+
self.scores = len(self.dataset_df["score"].values) * [1]
|
98 |
+
|
99 |
+
def __getitem__(self, query_idx):
|
100 |
+
|
101 |
+
protein_seq = self.protein_seqs[query_idx]
|
102 |
+
disease_des = self.disease_dess[query_idx]
|
103 |
+
score = self.scores[query_idx]
|
104 |
+
|
105 |
+
return protein_seq, disease_des, score
|
106 |
+
|
107 |
+
def __len__(self):
|
108 |
+
return len(self.protein_seqs)
|
109 |
+
# # 分离正负样本
|
110 |
+
# positive_samples = self.dataset_df[self.dataset_df["score"] == 1]
|
111 |
+
# negative_samples = self.dataset_df[self.dataset_df["score"] == 0]
|
112 |
+
|
113 |
+
# # 打乱并划分正样本
|
114 |
+
# positive_samples = positive_samples.sample(frac=1, random_state=42).reset_index(drop=True)
|
115 |
+
# num_pos_val_samples = int(len(positive_samples) * val_ratio)
|
116 |
+
|
117 |
+
# # 打乱并划分负样本
|
118 |
+
# negative_samples = negative_samples.sample(frac=1, random_state=42).reset_index(drop=True)
|
119 |
+
# num_neg_val_samples = int(len(negative_samples) * val_ratio)
|
120 |
+
|
121 |
+
# if split == "train":
|
122 |
+
# self.dataset_df = pd.concat([positive_samples[:-num_pos_val_samples], negative_samples[:-num_neg_val_samples]])
|
123 |
+
# print(f"{data_dir}disgenet_gda.csv loaded, found associations: {len(self.dataset_df.index)}")
|
124 |
+
# elif split == "val":
|
125 |
+
# self.dataset_df = pd.concat([positive_samples[-num_pos_val_samples:], negative_samples[-num_neg_val_samples:]])
|
126 |
+
# print(f"{data_dir}disgenet_gda.csv loaded, found associations: {len(self.dataset_df.index)}")
|
127 |
+
# Shuffle and split data
|
128 |
+
|
129 |
+
# class GDA_Pretrain_Dataset(Dataset):
|
130 |
+
# """
|
131 |
+
# Candidate Dataset for:
|
132 |
+
# ALL gene-disease associations
|
133 |
+
# """
|
134 |
+
|
135 |
+
# def __init__(self, data_dir="../../data/pretrain/", test=False):
|
136 |
+
# LOGGER.info("Initializing GDA Pretraining Dataset ! ...")
|
137 |
+
# updated = pd.read_csv(f"{data_dir}/disgenet_updated.csv")
|
138 |
+
|
139 |
+
# data = GDA(name="DisGeNET")
|
140 |
+
# data = data.get_data()
|
141 |
+
# data = data[['Gene_ID','Disease_ID']].dropna()
|
142 |
+
# self.dataset_df = pd.read_csv(f"{data_dir}/disgenet_gda.csv")
|
143 |
+
|
144 |
+
# num_unique_diseaseId = self.dataset_df['diseaseId'].nunique()
|
145 |
+
# num_unique_geneId = self.dataset_df['geneId'].nunique()
|
146 |
+
|
147 |
+
# print(f"Number of unique 'diseaseId': {num_unique_diseaseId}")
|
148 |
+
# print(f"Number of unique 'geneId': {num_unique_geneId}")
|
149 |
+
|
150 |
+
# num_of_c0002395 = self.dataset_df[self.dataset_df['diseaseId'] == 'C0002395'].shape[0]
|
151 |
+
# print(f"Alzheimer Number in 2020:{num_of_c0002395}")
|
152 |
+
|
153 |
+
# Convert 'Gene_ID' and 'Disease_ID' to str before merge
|
154 |
+
# data['Gene_ID'] = data['Gene_ID'].astype(str)
|
155 |
+
# data['Disease_ID'] = data['Disease_ID'].astype(str)
|
156 |
+
|
157 |
+
# Similarly for 'geneId' and 'diseaseId', if they're not already of type 'str'
|
158 |
+
# self.dataset_df['geneId'] = self.dataset_df['geneId'].astype(str)
|
159 |
+
# self.dataset_df['diseaseId'] = self.dataset_df['diseaseId'].astype(str)
|
160 |
+
|
161 |
+
# # 合并两个DataFrame并找出不同的行
|
162 |
+
# merged = df.merge(self.dataset_df, how='outer', indicator=True)
|
163 |
+
# differences = merged[merged['_merge'] != 'both']
|
164 |
+
|
165 |
+
# differences.to_csv('/nfs/dpa_pretrain/data/pretrain/differences.csv', index=False)
|
166 |
+
|
167 |
+
|
168 |
+
# Check for overlap between TDC dataset and DisGeNET dataset
|
169 |
+
# merged_df = pd.merge(data, self.dataset_df, how='inner', left_on=['Gene_ID','Disease_ID'], right_on=['geneId','diseaseId'])
|
170 |
+
|
171 |
+
# num_matched_pairs = merged_df.shape[0]
|
172 |
+
|
173 |
+
# print(f"Number of matched pairs TDC: {num_matched_pairs}")
|
174 |
+
|
175 |
+
# merged_dis = pd.merge(data, updated, how='inner', left_on=['Gene','Disease'], right_on=['proteinSeq','diseaseDes'])
|
176 |
+
|
177 |
+
# num_matched = merged_dis.shape[0]
|
178 |
+
|
179 |
+
# print(f"Number of matched pairs DisGeNET_test: {num_matched}")
|
180 |
+
|
181 |
+
# self.dataset_df = self.dataset_df[
|
182 |
+
# ["proteinSeq", "diseaseDes", "score"]
|
183 |
+
# ].dropna() # Drop missing values.
|
184 |
+
# print(self.dataset_df.head()) "proteinSeq", "diseaseDes", "score"
|
185 |
+
|
186 |
+
# print(
|
187 |
+
# f"{data_dir}disgenet_gda.csv loaded, found associations: {len(self.dataset_df.index)}"
|
188 |
+
# )
|
189 |
+
# df1 = pd.read_csv(f"{data_dir}/disgenet_gda.csv")
|
190 |
+
# df1 = df1[
|
191 |
+
# ["proteinSeq", "diseaseDes", "score"]
|
192 |
+
# ].dropna()
|
193 |
+
|
194 |
+
# # 合并两个DataFrame并找出不同的行
|
195 |
+
# merged = df1.merge(self.dataset_df, how='outer', indicator=True)
|
196 |
+
# differences = merged[merged['_merge'] != 'both']
|
197 |
+
|
198 |
+
# # 将结果保存到新的文件中
|
199 |
+
# differences.to_csv('/nfs/dpa_pretrain/data/pretrain/differences.csv', index=False)
|
200 |
+
|
201 |
+
# if test:
|
202 |
+
# self.protein_seqs = self.dataset_df["proteinSeq"].values[:128]
|
203 |
+
# self.disease_dess = self.dataset_df["diseaseDes"].values[:128]
|
204 |
+
# self.scores = 128 * [1]
|
205 |
+
# else:
|
206 |
+
# self.protein_seqs = self.dataset_df["proteinSeq"].values
|
207 |
+
# self.disease_dess = self.dataset_df["diseaseDes"].values
|
208 |
+
# self.scores = len(self.dataset_df["score"].values) * [1]
|
209 |
+
|
210 |
+
# def __getitem__(self, query_idx):
|
211 |
+
|
212 |
+
# protein_seq = self.protein_seqs[query_idx]
|
213 |
+
# disease_des = self.disease_dess[query_idx]
|
214 |
+
# score = self.scores[query_idx]
|
215 |
+
|
216 |
+
# return protein_seq, disease_des, score
|
217 |
+
|
218 |
+
# def __len__(self):
|
219 |
+
# return len(self.protein_seqs)
|
220 |
+
|
221 |
+
|
222 |
+
class PPI_Pretrain_Dataset(Dataset):
|
223 |
+
"""
|
224 |
+
Candidate Dataset for:
|
225 |
+
ALL protein-to-protein interactions
|
226 |
+
"""
|
227 |
+
|
228 |
+
def __init__(self, data_dir="../../data/pretrain/", test=False):
|
229 |
+
LOGGER.info("Initializing metric learning data set! ...")
|
230 |
+
self.dataset_df = pd.read_csv(f"{data_dir}/string_ppi_900_2m.csv")
|
231 |
+
self.dataset_df = self.dataset_df[["item_seq_a", "item_seq_b", "score"]]
|
232 |
+
self.dataset_df = self.dataset_df.dropna()
|
233 |
+
if test:
|
234 |
+
self.dataset_df = self.dataset_df.sample(100)
|
235 |
+
print(
|
236 |
+
f"{data_dir}/string_ppi_900_2m.csv loaded, found interactions: {len(self.dataset_df.index)}"
|
237 |
+
)
|
238 |
+
self.protein_seq1 = self.dataset_df["item_seq_a"].values
|
239 |
+
self.protein_seq2 = self.dataset_df["item_seq_b"].values
|
240 |
+
self.scores = len(self.dataset_df["score"].values) * [1]
|
241 |
+
|
242 |
+
def __getitem__(self, query_idx):
|
243 |
+
|
244 |
+
protein_seq1 = self.protein_seq1[query_idx]
|
245 |
+
protein_seq2 = self.protein_seq2[query_idx]
|
246 |
+
score = self.scores[query_idx]
|
247 |
+
|
248 |
+
return protein_seq1, protein_seq2, score
|
249 |
+
|
250 |
+
def __len__(self):
|
251 |
+
return len(self.protein_seq1)
|
252 |
+
|
253 |
+
|
254 |
+
class PPI_Dataset(Dataset):
|
255 |
+
"""
|
256 |
+
Candidate Dataset for:
|
257 |
+
ALL protein-to-protein interactions
|
258 |
+
"""
|
259 |
+
|
260 |
+
def __init__(self, protein_seq1, protein_seq2, score):
|
261 |
+
self.protein_seq1 = protein_seq1
|
262 |
+
self.protein_seq2 = protein_seq2
|
263 |
+
self.scores = score
|
264 |
+
|
265 |
+
def __getitem__(self, query_idx):
|
266 |
+
|
267 |
+
protein_seq1 = self.protein_seq1[query_idx]
|
268 |
+
protein_seq2 = self.protein_seq2[query_idx]
|
269 |
+
score = self.scores[query_idx]
|
270 |
+
|
271 |
+
return protein_seq1, protein_seq2, score
|
272 |
+
|
273 |
+
def __len__(self):
|
274 |
+
return len(self.protein_seq1)
|
275 |
+
|
276 |
+
|
277 |
+
class DDA_Dataset(Dataset):
|
278 |
+
"""
|
279 |
+
Candidate Dataset for:
|
280 |
+
ALL disease-to-disease associations
|
281 |
+
"""
|
282 |
+
|
283 |
+
def __init__(self, diseaseDes1, diseaseDes2, label):
|
284 |
+
self.diseaseDes1 = diseaseDes1
|
285 |
+
self.diseaseDes2 = diseaseDes2
|
286 |
+
self.label = label
|
287 |
+
|
288 |
+
def __getitem__(self, query_idx):
|
289 |
+
|
290 |
+
diseaseDes1 = self.diseaseDes1[query_idx]
|
291 |
+
diseaseDes2 = self.diseaseDes2[query_idx]
|
292 |
+
label = self.label[query_idx]
|
293 |
+
|
294 |
+
return diseaseDes1, diseaseDes2, label
|
295 |
+
|
296 |
+
def __len__(self):
|
297 |
+
return len(self.diseaseDes1)
|
298 |
+
|
299 |
+
|
300 |
+
class DDA_Pretrain_Dataset(Dataset):
|
301 |
+
"""
|
302 |
+
Candidate Dataset for:
|
303 |
+
ALL protein-to-protein interactions
|
304 |
+
"""
|
305 |
+
|
306 |
+
def __init__(self, data_dir="../../data/pretrain/", test=False):
|
307 |
+
LOGGER.info("Initializing metric learning data set! ...")
|
308 |
+
self.dataset_df = pd.read_csv(f"{data_dir}disgenet_dda.csv")
|
309 |
+
self.dataset_df = self.dataset_df.dropna() # Drop missing values.
|
310 |
+
if test:
|
311 |
+
self.dataset_df = self.dataset_df.sample(100)
|
312 |
+
print(
|
313 |
+
f"{data_dir}disgenet_dda.csv loaded, found associations: {len(self.dataset_df.index)}"
|
314 |
+
)
|
315 |
+
self.disease_des1 = self.dataset_df["diseaseDes1"].values
|
316 |
+
self.disease_des2 = self.dataset_df["diseaseDes2"].values
|
317 |
+
self.scores = len(self.dataset_df["jaccard_variant"].values) * [1]
|
318 |
+
|
319 |
+
def __getitem__(self, query_idx):
|
320 |
+
|
321 |
+
disease_des1 = self.disease_des1[query_idx]
|
322 |
+
disease_des2 = self.disease_des2[query_idx]
|
323 |
+
score = self.scores[query_idx]
|
324 |
+
|
325 |
+
return disease_des1, disease_des2, score
|
326 |
+
|
327 |
+
def __len__(self):
|
328 |
+
return len(self.disease_des1)
|
src/utils/downstream_disgenet.py
CHANGED
@@ -11,9 +11,9 @@ import pandas as pd
|
|
11 |
sys.path.append("../")
|
12 |
|
13 |
class DisGeNETProcessor:
|
14 |
-
def __init__(self,input_csv_path
|
15 |
-
train_data = pd.read_csv('
|
16 |
-
valid_data = pd.read_csv('
|
17 |
test_data = pd.read_csv(input_csv_path)
|
18 |
|
19 |
# test_data = pd.read_csv('/nfs/dpa_pretrain/data/downstream/GDA_Data/test.csv')
|
|
|
11 |
sys.path.append("../")
|
12 |
|
13 |
class DisGeNETProcessor:
|
14 |
+
def __init__(self,input_csv_path):
|
15 |
+
train_data = pd.read_csv('data/downstream/GDA_Data/train.csv')
|
16 |
+
valid_data = pd.read_csv('data/downstream/GDA_Data/valid.csv')
|
17 |
test_data = pd.read_csv(input_csv_path)
|
18 |
|
19 |
# test_data = pd.read_csv('/nfs/dpa_pretrain/data/downstream/GDA_Data/test.csv')
|
src/utils/gd_model.py
ADDED
@@ -0,0 +1,77 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import sys
|
2 |
+
|
3 |
+
import torch
|
4 |
+
import torch.nn as nn
|
5 |
+
|
6 |
+
sys.path.append("../")
|
7 |
+
|
8 |
+
|
9 |
+
class GDANet(torch.nn.Module):
|
10 |
+
def __init__(
|
11 |
+
self,
|
12 |
+
prot_encoder,
|
13 |
+
disease_encoder,
|
14 |
+
):
|
15 |
+
"""_summary_
|
16 |
+
|
17 |
+
Args:
|
18 |
+
prot_encoder (_type_): _description_
|
19 |
+
disease_encoder (_type_): _description_
|
20 |
+
prot_out_dim (int, optional): _description_. Defaults to 1024.
|
21 |
+
disease_out_dim (int, optional): _description_. Defaults to 768.
|
22 |
+
drop_out (int, optional): _description_. Defaults to 0.
|
23 |
+
freeze_prot_encoder (bool, optional): _description_. Defaults to True.
|
24 |
+
freeze_disease_encoder (bool, optional): _description_. Defaults to True.
|
25 |
+
"""
|
26 |
+
super(GDANet, self).__init__()
|
27 |
+
self.prot_encoder = prot_encoder
|
28 |
+
self.disease_encoder = disease_encoder
|
29 |
+
self.cls = None
|
30 |
+
self.reg = None
|
31 |
+
|
32 |
+
def add_regression_head(self, prot_out_dim=1024, disease_out_dim=768):
|
33 |
+
"""Add regression head.
|
34 |
+
|
35 |
+
Args:
|
36 |
+
prot_out_dim (_type_): protein encoder output dimension.
|
37 |
+
disease_out_dim (_type_): disease encoder output dimension.
|
38 |
+
drop_out (int, optional): dropout rate. Defaults to 0.
|
39 |
+
"""
|
40 |
+
self.reg = nn.Linear(prot_out_dim + disease_out_dim, 1)
|
41 |
+
|
42 |
+
|
43 |
+
def add_classification_head(
|
44 |
+
self, prot_out_dim=1024, disease_out_dim=768, out_dim=2
|
45 |
+
):
|
46 |
+
"""Add classification head.
|
47 |
+
|
48 |
+
Args:
|
49 |
+
prot_out_dim (_type_): protein encoder output dimension.
|
50 |
+
disease_out_dim (_type_): disease encoder output dimension.
|
51 |
+
out_dim (int, optional): output dimension. Defaults to 2.
|
52 |
+
drop_out (int, optional): dropout rate. Defaults to 0.
|
53 |
+
"""
|
54 |
+
self.cls = nn.Linear(prot_out_dim + disease_out_dim, out_dim)
|
55 |
+
|
56 |
+
|
57 |
+
def freeze_encoders(self, freeze_prot_encoder, freeze_disease_encoder):
|
58 |
+
"""Freeze encoders.
|
59 |
+
|
60 |
+
Args:
|
61 |
+
freeze_prot_encoder (boolean): freeze protein encoder
|
62 |
+
freeze_disease_encoder (boolean): freeze disease textual encoder
|
63 |
+
"""
|
64 |
+
if freeze_prot_encoder:
|
65 |
+
for param in self.prot_encoder.parameters():
|
66 |
+
param.requires_grad = False
|
67 |
+
else:
|
68 |
+
for param in self.disease_encoder.parameters():
|
69 |
+
param.requires_grad = True
|
70 |
+
if freeze_disease_encoder:
|
71 |
+
for param in self.disease_encoder.parameters():
|
72 |
+
param.requires_grad = False
|
73 |
+
else:
|
74 |
+
for param in self.disease_encoder.parameters():
|
75 |
+
param.requires_grad = True
|
76 |
+
print(f"freeze_prot_encoder:{freeze_prot_encoder}")
|
77 |
+
print(f"freeze_disease_encoder:{freeze_disease_encoder}")
|