blob: 9a6c317df13ddca1a4e25c71f28b5adb77713bba (
plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
|
# Curie-Web
Tested on:
* macOS 10.15 (Catalina)
* Ubuntu 20.04 - Raspberry Pi 4
# Quick Start (Docker-Compose)
You can quickly get started and test Curie-Web without needing to manage dependencies by using the `docker-compose` image. This has all features except AR/3D Model support. Otherwise, it is production ready.
```
git clone https://github.com/navanchauhan/Curie-Web
cd Curie-Web
docker-compose up
```
Do not forget to edit `config.ini` for the email section. Do not change the database settings for running it via docker.
## Caveat(s):
* This does not support generating a 3D Model and thus, AR viewer will not work on the Job Status page.
# Installation
## 1. Installing Dependencies
### 1.1 Docker
Once you have installed docker, make sure to pull the following images (Otherwise, these will automatically get downloaded when you run the web-server)
* navanchauhan/curie-cli (amd64/aarch64)
* navanchauhan/usd-from-gltf (aarch64)
* leon/usd-from-gltf (amd64)
### 1.2 PLIP
Install from [pharmai/plip](https://github.com/pharmai/plip).
In case you have problems installing it, install it from the forked repo [navanchauhan/plip](https://github.com/navanchauhan/plip)
### 1.3 PyMOL with Python Bindings (version >= 2.0)
* macOS users can use Homebrew to install it via `brew install pymol`
* Users using apt can install it via `sudo apt install pymol`
### 1.4 Open-Babel (version >= 3.0)
macOS users can use Homebrew to install it via `brew install open-babel`
Users using apt can install it via `sudo apt install openbabel python3-openbabel`
### 1.5 Tensorflow (Optional)
Please ensure you install Tensorflow 2.x.
If tensorflow is not properly installed, it will automatically disable the LSTM Module.
## 2. Changing the Configuration
Replace the values in `config.ini`
Default Values:
```ini
[DATABASE]
HOST = navanspi.duckdns.org
PORT = 3306
USER = curieweb
PASSWORD = curie-web-russian-54
NAME = curie
[SMTP]
SERVER = smtp.gmail.com
PORT = 587
EMAIL = navanchauhan@gmail.com
PASSWORD = okrs shoc ahtk idui
[LOGS]
LOG = True
SAVE_LOGS = False
[FILES]
UPLOAD_FOLDER = ./app/static/uploads
LOG_FOLDER = ./app/logs/
[EXECUTION]
INSTANT = True
[FEATURES]
LSTM = False
```
### **Database**
| Name | Description |
|----------|---------------------|
| HOST | MySQL Database Host |
| PORT | MySQL Database Port |
| USER | Username |
| PASSWORD | Password |
| NAME | MySQL Database Name |
### **SMTP**
| Name | Description |
|----------|-------------|
| SERVER | SMTP Server |
| PORT | SMTP Port |
| EMAIL | Email |
| PASSWORD | Password |
### **LOGS**
| Name | Description |
|-----------|-------------|
| LOG | Log |
| SAVE_LOGS | SAVE LOGS |
### **FILES**
| Name | Description |
|---------------|-----------------------|
| UPLOAD_FOLDER | Folder to store files |
| LOG_FOLDER | Folder to store logs |
### **EXECUTION**
| Name | Description |
|---------------|-----------------------|
| INSTANT | Whether to run the docking jobs instantly (True or False) |
### **FEATURES**
| Name | Description |
|---------------|-----------------------|
| LSTM | Enable LSTM Generator (True or False) |
## 3. Adding AR Model Support
Make sure you have PyMOL 2.0 or higher
Either download the precompiled binaries from [COLLADA2GLTF](https://github.com/KhronosGroup/COLLADA2GLTF) or compile it on your own
Once you have the `COLLADA2GLTF-bin` file, copy the file:
```
cp COLLADA2GLTF-bin /usr/local/bin/collada2gltf
```
## 4. Setting up the Database
You will first need to create a database and grant all priviliges to a user. Make sure you have correctly configured the `config.ini` file.
After tha simply run the following commands. This will create the table(s) and check if the backend is working or not.
```
cd tests
python3 dbTestFiller.py
python3 backendTest.py
python3 removeSample.py
```
## 5. Running
### Without API Support
`gunicorn app:app -b "0.0.0.0:7589"`
### With API Support (FastAPI)
`gunicorn api:app -k uvicorn.workers.UvicornWorker -b "0.0.0.0:7589"`
### systemd
**There is a sample systemd file in the misc folder**
* Configure the file and then copy it to `/etc/systemd/system`
* You can start the server by running `sudo systemctl start curie`
* To enable the server to start on boot run `sudo systemctl enable curie`
## Misc.
### Fine-Tuning the LSTM Module
Currently the base model has been trained on 1.5 million compounds and took ~4 hours to train on Google Colab's GPU runtime.
*Information to be added soon*
|
