jrwnter / gruenifai

Implementation grünif.ai: Interactive multi-parameter optimization of molecules in a continuous vector space

gruenif.ai

gruenif.ai is a web application for interactive multi-parameter optimization of molecules in a continous vector space.

Showcase of functionality

A web application is best explained through a live demo. Please watch our video for a detailed explanation of the functionality: Showcase

Installation and setting up

In its current form grünifai is commposed of multiple microservices. We recommend the use of Docker to compose all microservises into one running environment.

Installation using Docker

• clone this environment.
• cd <REPOSITORY_DIR>
• docker compose up

The web app can now be accessed at http://localhost:8777/ When running without GPUs or to test, we recommend to set the checkbox for fastmode to reduce compute time.

Installation using Anaconda

The repository can also be installed from scratch. That way GPU support can be enabled as well. However, currently, we recommend using the Docker installation described above.

Dependencies

• cddd
• mso
• psycopg2
• flask
• flask_restplus
• requests
• nodejs
• yarn
• pyzmq
• marvinjs for molecule sketching: https://chemaxon.com/products/marvin-js

Installation

1. terminal

• clone and install cddd
• clone and install mso
• clone this repository
• cd <REPOSITORY_DIR>
• conda env create -f environment.yml ( + activate this environment)
• pip install .
• cd <REPOSITORY_DIR>/gruenifai/backend/postgres
• python create_db.py (assumes a running postgres server running locally)
• cd <REPOSITORY_DIR>/gruenifai/gui/client
• yarn install
• yarn start
• cp marvinjs forlder into /gruenifai/gui/client/public

2. terminal

• cd <REPOSITORY_DIR>/gruenifai/gui/server
• python api.py

3. terminal

• cd <REPOSITORY_DIR>/gruenifai/backend/
• python start_inference_server.py --model_dir <PATH_TO_CDDD_MODEL> --device <CUDA_DEVICE (e.g. 1 3 4)> --num_servers <NUMBER_OF_SERVERS>

4. terminal

• cd <REPOSITORY_DIR>/gruenifai/backend/
• python flaskserver.py --num_swarms=<NUMBER_OF_SWARMS> --num_workers=<NUMBER_OF_WORKERS>

The web app can now be accessed at port 3000

gruenifai's Issues

yarn install

I get the error massage:
error An unexpected error occurred: "https://registry.yarnpkg.com/websocket-driver/-/websocket-driver-0.7.2.tgz: Request failed "404 Not Found"".

when doing "yarn install" @andreassteffen

failed to load reward_penalized_log_p score. Consider installing package networkx

I noticed this on the log of docker-compose:

backend_1      | /opt/conda/envs/cddd/lib/python3.6/site-packages/mso/objectives/mol_functions.py:211: UserWarning: failed to load reward_penalized_log_p score. Consider installing package networkx
backend_1      |   warnings.warn("failed to load reward_penalized_log_p score. Consider installing package networkx")

Does the lack of networkx break functionality?

Thanks,
João

export molecules fails with ModuleNotFoundError: No module named 'xlwt'

Hi,

I've followed the instructions for deploying using docker-compose. It was successfully deployed, I am able to use the interface, but when exporting the molecules from the "cart", it errors with:

Traceback (most recent call last):
  File "/opt/conda/envs/gfai/lib/python3.7/site-packages/flask/app.py", line 2463, in __call__
    return self.wsgi_app(environ, start_response)
  File "/opt/conda/envs/gfai/lib/python3.7/site-packages/flask/app.py", line 2449, in wsgi_app
    response = self.handle_exception(e)
  File "/opt/conda/envs/gfai/lib/python3.7/site-packages/flask_restplus/api.py", line 584, in error_router
    return original_handler(e)
  File "/opt/conda/envs/gfai/lib/python3.7/site-packages/flask/app.py", line 1866, in handle_exception
    reraise(exc_type, exc_value, tb)
  File "/opt/conda/envs/gfai/lib/python3.7/site-packages/flask/_compat.py", line 38, in reraise
    raise value.with_traceback(tb)
  File "/opt/conda/envs/gfai/lib/python3.7/site-packages/flask/app.py", line 2446, in wsgi_app
    response = self.full_dispatch_request()
  File "/opt/conda/envs/gfai/lib/python3.7/site-packages/flask/app.py", line 1951, in full_dispatch_request
    rv = self.handle_user_exception(e)
  File "/opt/conda/envs/gfai/lib/python3.7/site-packages/flask_restplus/api.py", line 584, in error_router
    return original_handler(e)
  File "/opt/conda/envs/gfai/lib/python3.7/site-packages/flask/app.py", line 1820, in handle_user_exception
    reraise(exc_type, exc_value, tb)
  File "/opt/conda/envs/gfai/lib/python3.7/site-packages/flask/_compat.py", line 38, in reraise
    raise value.with_traceback(tb)
  File "/opt/conda/envs/gfai/lib/python3.7/site-packages/flask/app.py", line 1949, in full_dispatch_request
    rv = self.dispatch_request()
  File "/opt/conda/envs/gfai/lib/python3.7/site-packages/flask/app.py", line 1935, in dispatch_request
    return self.view_functions[rule.endpoint](**req.view_args)
  File "/opt/conda/envs/gfai/lib/python3.7/site-packages/flask_restplus/api.py", line 325, in wrapper
    resp = resource(*args, **kwargs)
  File "/opt/conda/envs/gfai/lib/python3.7/site-packages/flask/views.py", line 89, in view
    return self.dispatch_request(*args, **kwargs)
  File "/opt/conda/envs/gfai/lib/python3.7/site-packages/flask_restplus/resource.py", line 44, in dispatch_request
    resp = meth(*args, **kwargs)
  File "/home/server/api.py", line 238, in post
    data.to_excel(fn, index=None)
  File "/opt/conda/envs/gfai/lib/python3.7/site-packages/pandas/core/generic.py", line 2181, in to_excel
    engine=engine,
  File "/opt/conda/envs/gfai/lib/python3.7/site-packages/pandas/io/formats/excel.py", line 726, in write
    writer = ExcelWriter(stringify_path(writer), engine=engine)
  File "/opt/conda/envs/gfai/lib/python3.7/site-packages/pandas/io/excel/_xlwt.py", line 13, in __init__
    import xlwt
ModuleNotFoundError: No module named 'xlwt'

I quickly checked and apparently xlwt is not a pandas dependency, so I guess it needs to be added explicitly to the environment.yml of the gui server.

At least what I did was add xlwt to https://github.com/jrwnter/gruenifai/blob/master/gruenifai/gui/server/environment.yml, build that image locally instead of pulling it, and the export works.

I hope this is helpful.
João