Chemical Property Prediction

Introduction

This guide will go over how to make use of chemical property prediction workflow using the API. You will need to make sure you are importing the OSlibrary into your python code.

Finding The Module

The first thing you need to find is the id for the chemical prediction module, which can be found at the /layar/module.

moduleUrl = f'{envUrl}/layar/module'
moduleId = ""

header = {'Accept': 'application/json',
          'Content-Type': 'application/json',
	  'X-Vyasa-Data-Providers': 'internal-prod.certara.ai',
          'Authorization': f"Bearer {token}",
          'X-Vyasa-Data-Fabric': fabricId}


response = requests.get(moduleUrl, headers = header).json()

for i in response:
  if i.get("name") == 'Molecular Property Prediction':
    moduleId = i.get("id")
  else:
    continue

Create The Project

With the moduleId we can create the project using the /layar/project endpoint.

projectUrl = f'{envUrl}/layar/project'

body = {
'moduleId': moduleId,
'name': 'test predict',
'description': 'test predict'}

response = requests.post(projectUrl, headers = header, json = body).json()

projectId = response.get("id")

Start The Project

Start the project using the projectId and the /layar/project/projectId/runendpoint. We can then use the /layar/projectComputation/projectIdendpoint to check the status.

projectRunUrl = f'{envUrl}/layar/project/{projectId}/run'

body = {
  'computationParameters': {
    'smiles_raw': '<SMILE strings seperated by new lines>'},
  'moduleId': moduleId 
}

response = requests.post(projectRunUrl, headers = header, json = body).json()

status = None
    while status != 'complete':
        response = requests.get(f'{envUrl}/layar/projectComputation/{projectId}',
                              headers = header
                             ).json().get("status")
        if response == 'Complete!':
            status = 'complete'
            print(id + ' Done running')
            return id
        else:
            status = None

Download Results

You can then download the results of the prediction using the /layar/projectComputation/projectId/downloadendpoint.

downloadProjectUrl = f'{envUrl}/layar/projectComputation/{projectId}/download?access_token={token}'
placeFile = 'C:\users\test\downloads\'

#functions to extract filename from request headers and make sure they are in an acceptable format
def getFileName(cd):
    fname = re.findall('filename=(.+)', cd)
    fname = fname[0]
    fname = fname.replace('"', '')
    fname = fname.replace(';', '')
    return fname

def sanitizeName(name):
    length = len(name)
    while length > 150:
        name = name[:140]
        length = len(name)
    return name

response = requests.get(downloadProjectUrl,
                        headers = {'Accept': 'application/json'},
                        stream = True)

name = getFileName(response.headers.get('content-disposition'))
name = sanitizeName(name)

with open(f'{placeFile}/{name}.csv', 'wb') as file:
	for chunk in response.iter_content(chunk_size=8192):
            file.write(chunk)