Upload 500+ materials to CRIPT from Excel file

[nh916]

import logging
import os
import uuid
from pathlib import Path

import pandas as pd

import cript

# read data from Excel file using Pandas
path_to_excel_file = Path("./Validation.xlsx").resolve()

# read Excel file into pandas df
df = pd.read_excel(
    io=path_to_excel_file, sheet_name="Small Molecule Queries", header=[0, 1, 2, 3]
)


# create a log file and log errors to file in same directory
logging.basicConfig(filename='error_log.txt', level=logging.ERROR, format='%(message)s')


# create API client
with cript.API(
        host=host,
        api_token=api_token,
        storage_token=os.getenv("CRIPT_STORAGE_TOKEN"),
) as api:

    # turn off api logs in terminal
    api.verbose = False

    # create new project
    # use `uuid` at the end of the project name to avoid duplicate project error (THIS IS FOR TESTING PURPOSES ONLY)
    my_project = cript.Project(name=f"material_search_test_project__{uuid.uuid4().hex}")

    my_project.material = []

    print(f"\n\n PROJECT NAME: {my_project.name}")
    print(f" PROJECT UUID: {my_project.uuid} \n\n")

    # loop over rows in Excel sheet, and create materials with it
    for index, row in df.iterrows():
        name_row = row["Name"].values[0]

        # if name_row is empty, then use the default value, if it has a value then use the value
        if not pd.notna(name_row):
            # adding 5 to index to get the correct row number from Excel since we are skipping the first 5 rows of the excel file
            name_row = f"row: {index + 5}"

        # use `uuid` at the end of the material name to avoid duplicate material error (THIS IS FOR TESTING PURPOSES ONLY)
        material_name = f"{name_row}__{uuid.uuid4().hex}"

        # get bigsmiles value from Excel file from pandas series
        bigsmiles = row["Targets (BigSMILES)"].values[0]

        print(f"Name: `{material_name}`    BigSMILES: `{bigsmiles}`")

        # create a new material and set its identifier
        new_material = cript.Material(
            name=material_name, identifiers=[{"bigsmiles": bigsmiles}]
        )

        # attach each new material to project
        my_project.material += [new_material]

        # show developers what row the program is on
        print(f"on row: {index + 5} \n")

        try:
            api.save(project=my_project)
        except Exception as error:
            # log errors to a log file and continue
            logging.exception(f"Material name: {new_material.name} || row: {index + 5}     |||   {error}")

            print(f"error || Material name: {new_material.name} || row: {index + 5}")
            print(error)

       # remove already saved nodes on every 10th iteration because API has a limit of only 100 nodes for each request
        if index % 10 == 0:
            my_project.material = []

    my_paginator = api.search(node_type=cript.Project, search_mode=cript.SearchModes.EXACT_NAME,
                              value_to_search=my_project.name)

    # get the project from paginator
    my_project_from_api_dict = my_paginator.current_page_results[0]

    # after save generate the website URL for me to easily see it online
    print(f"https://www.mycriptapp.org/project/{my_project_from_api_dict['uuid']}")

[nh916]

A great script that shows how to reuse already existing material nodes from the CRIPT DB in your project, by @InnocentBug

import os
import json
import cript
from cript.api.exceptions import CRIPTAPISaveError

uniq = "9"

data = [
    {"name": "askdfjasdf"+uniq, "bigsmiles": "NXC"},
    {"name": "Ammonia", "bigsmiles": "N"},
    {"name": "benzene", "bigsmiles": "c1ccccc1"},
    {"name": "askdfja"+uniq, "bigsmiles": "NXC"},
]


# create API client
with cript.API(
    host=os.getenv("CRIPT_HOST"),
    api_token=os.getenv("CRIPT_TOKEN"),
    storage_token=os.getenv("CRIPT_STORAGE_TOKEN"),
) as api:
    # create new project
    my_project = cript.Project(name="material_search_test_project" + uniq)

    # loop over rows in Excel sheet, and create materials with it
    for i, d in enumerate(data):
        name = d["name"]
        # create a new material and set its identifier
        new_material = cript.Material(name=name, identifiers=[{"bigsmiles": d["bigsmiles"]}])

        my_paginator = api.search(node_type=cript.Material, search_mode=cript.SearchModes.EXACT_NAME, value_to_search=new_material.name)
        if len(my_paginator.current_page_results) > 0:
            my_material_from_api_dict = my_paginator.current_page_results[0]
            try:
                new_material = cript.load_nodes_from_json(json.dumps(my_material_from_api_dict))
            except Exception as exc:
                print(exc)
            else:
                print("reuse material", new_material.name)
                # attach each new material to project
                my_project.material += [new_material]
        else:
            my_project.material += [new_material]

        api.save(project=my_project)

    print(my_project.get_json(indent=2).json)
    my_paginator = api.search(node_type=cript.Project, search_mode=cript.SearchModes.EXACT_NAME, value_to_search=my_project.name)

    # get the project from paginator
    my_project_from_api_dict = my_paginator.current_page_results[0]
    print(json.dumps(my_project_from_api_dict, indent=2))

[nh916]

Please note, this script is designed to add material nodes to a project in the CRIPT database one at a time. While it's a handy utility for quickly populating your project with materials, it does not establish connections between the nodes as they are saved in chunks.

This means that if you are looking for a way to create a fully-connected, complex network of material nodes, this script may not be suitable. It's intended for more straightforward use-cases where you simply want to add a list of materials to a project without linking them together.