caddiepy 0.3.0

The python package to the Cancer Driver Drug Interaction Explorer (CADDIE)

caddiepy

The python package to the Cancer Driver Drug Interaction Explorer (CADDIE). It provides an interface for a variety of CADDIEs functionalities, giving the user the possibility to execute tasks on CADDIE programmatically without using the website. This allows to run a larger number of drug-target search or drug repurposing tasks and to implement CADDIE into your programs. For more information about CADDIE, visit the documentation.

Working example

import caddiepy

gene_list = ['PTEN', 'MYC']

# verify that genes exist in CADDIE
caddie_gene_list = caddiepy.api.map_gene_id(gene_list).json()['genes']

task = caddiepy.Task('drug', 'network_proximity', [gene['graphId'] for gene in caddie_gene_list])
# customize parameters
task.set_parameter('result_size', 5)
task.set_parameter('include_indirect_drugs', True)
task.set_parameter('gene_interaction_datasets', ['IID'])
task.set_parameter('drug_interaction_datasets', ['DrugBank'])
# start task
task.run()
# get result when task is finished, if task is still running wait until it is finished
result = task.get_result()

Pipeline example (drug target search folled by drug search)

Note: The following pipeline can also be called as:

gene_list = ['FBXW7', 'PTEN', 'MYC']
resulting_drugs = caddiepy.pipeline(gene_list)['drugObjects']

gene_list = ['FBXW7', 'PTEN', 'MYC']

# verify that genes exist in CADDIE
caddie_gene_list = caddiepy.api.map_gene_id(gene_list).json()['genes']

task = caddiepy.Task('drug-target', 'multi-steiner', [gene['graphId'] for gene in caddie_gene_list])
# customize parameters
task.set_parameter('result_size', 20)
task.set_parameter('gene_interaction_datasets', ['BioGRID'])
# start task
task.run()
# get result when task is finished, if task is still running wait until it is finished
result_drug_targets = task.get_result()

task = caddiepy.Task('drug', 'trustrank', result_drug_targets['network']['nodes'])
# customize parameters
task.set_parameter('result_size', 20)
task.set_parameter('include_indirect_drugs', True)
task.set_parameter('gene_interaction_datasets', ['IID'])
task.set_parameter('drug_interaction_datasets', ['DrugBank'])
# start task
task.run()
# get result when task is finished, if task is still running wait until it is finished
resulting_drugs = task.get_result()['drugObjects']

How to use

Import

Import the module:

import caddiepy

2 Steps to repurpose drugs and find drug-targets

Step 1: Map genes to CADDIE gene IDs. This step is necessary to verify the genes exist in the CADDIE database. The gene objects returned contains a mapping to different IDs.

# gene list is a list of gene identifiers: entrez, uniprot ac or hugo
caddiepy.api.map_gene_id(gene_list)

# example to receive list of caddie IDs
caddie_gene_list = caddiepy.api.map_gene_id(gene_list).json()['genes']
caddie_gene_id_list = [gene['graphId'] for gene in caddie_gene_list]

Step 2: Use CADDIE IDs to find putative drug-targets or candidate drugs using one of CADDIEs algorithms

Drug-target algorithms:

• multisteiner
• keypathwayminer
• trustrank
• harmonic_centraliy
• degree_centraliy
• betweenness_centraliy

Drug algorithms:

• trustrank
• harmonic_centraliy
• degree_centraliy
• network_proximity

# target is either 'drug' or 'drug-target'
# caddie_gene_id_list is a list of caddie gene IDs (like g123, g234, ...)
task = caddiepy.Task(target, algorithm, caddie_gene_id_list)
# set parameters like this (for a full list of parameters and the available datasets look below)
task.set_parameter('result_size', 50)
task.set_parameter('gene_interaction_datasets', [gene_interaction_dataset1, gene_interaction_dataset2, ...])
task.set_parameter('drug_interaction_datasets', [drug_interaction_dataset1, drug_interaction_dataset2, ...])
# start task
task.run()
# get result when task is finished, if task is still running wait until it is finished
result = task.get_result()

Algorithm parameters

The full list of parameters for each algorithm (for an explanation, visit the documentation). For all available input options (dataset names, cancer-types) see below.

# multisteiner
task.set_parameter('num_trees', 5)
task.set_parameter('tolerance', 10)
task.set_parameter('hub_penalty', 0)
task.set_parameter('max_deg', sys.maxsize)
task.set_parameter('gene_interaction_datasets', gene_interaction_dataset_list)
task.set_parameter('mutation_cancer_type', mutation_cancer_type)
task.set_parameter('expression_cancer_type', expression_cancer_type)
task.set_parameter('tissue', tissue)

# keypathwayminer
task.set_parameter('k', 5)

# trustrank
task.set_parameter('damping_factor', 0.85)
task.set_parameter('hub_penalty', 0)
task.set_parameter('max_deg', sys.maxsize)
task.set_parameter('gene_interaction_datasets', gene_interaction_dataset_list)
task.set_parameter('drug_interaction_datasets', drug_interaction_dataset_list)
task.set_parameter('mutation_cancer_type', mutation_cancer_type)
task.set_parameter('expression_cancer_type', expression_cancer_type)
task.set_parameter('tissue', tissue)
task.set_parameter('include_nutraceutical_drugs', boolean)
task.set_parameter('only_atc_l_drugs', boolean)
task.set_parameter('include_indirect_drugs', boolean)
task.set_parameter('include_non_approved_drugs', boolean)
task.set_parameter('filter_paths', boolean)
task.set_parameter('available_drugs', available_drug_list)
# only avaibale if drug_interaction_dataset_list = ['drugbank']
task.set_parameter('drug_target_action', drug_effect)

# degree_centrality
task.set_parameter('hub_penalty', 0)
task.set_parameter('max_deg', sys.maxsize)
task.set_parameter('gene_interaction_datasets', gene_interaction_dataset_list)
task.set_parameter('drug_interaction_datasets', drug_interaction_dataset_list)
task.set_parameter('mutation_cancer_type', mutation_cancer_type)
task.set_parameter('expression_cancer_type', expression_cancer_type)
task.set_parameter('tissue', tissue)
task.set_parameter('include_nutraceutical_drugs', boolean)
task.set_parameter('only_atc_l_drugs', boolean)
task.set_parameter('include_indirect_drugs', boolean)
task.set_parameter('include_non_approved_drugs', boolean)
task.set_parameter('filter_paths', boolean)
task.set_parameter('available_drugs', available_drug_list)
# only avaibale if drug_interaction_dataset_list = ['drugbank']
task.set_parameter('drug_target_action', drug_effect)

# harmonic_centrality
task.set_parameter('hub_penalty', 0)
task.set_parameter('max_deg', sys.maxsize)
task.set_parameter('gene_interaction_datasets', gene_interaction_dataset_list)
task.set_parameter('drug_interaction_datasets', drug_interaction_dataset_list)
task.set_parameter('mutation_cancer_type', mutation_cancer_type)
task.set_parameter('expression_cancer_type', expression_cancer_type)
task.set_parameter('tissue', tissue)
task.set_parameter('include_nutraceutical_drugs', boolean)
task.set_parameter('only_atc_l_drugs', boolean)
task.set_parameter('include_indirect_drugs', boolean)
task.set_parameter('include_non_approved_drugs', boolean)
task.set_parameter('filter_paths', boolean)
task.set_parameter('available_drugs', available_drug_list)
# only avaibale if drug_interaction_dataset_list = ['drugbank']
task.set_parameter('drug_target_action', drug_effect)

# betweenness_centrality
task.set_parameter('hub_penalty', 0)
task.set_parameter('max_deg', sys.maxsize)
task.set_parameter('gene_interaction_datasets', gene_interaction_dataset_list)
task.set_parameter('mutation_cancer_type', mutation_cancer_type)
task.set_parameter('expression_cancer_type', expression_cancer_type)
task.set_parameter('tissue', tissue)
task.set_parameter('filter_paths', boolean)

# network_proximity
task.set_parameter('hub_penalty', 0)
task.set_parameter('max_deg', sys.maxsize)
task.set_parameter('gene_interaction_datasets', gene_interaction_dataset_list)
task.set_parameter('drug_interaction_datasets', drug_interaction_dataset_list)
task.set_parameter('mutation_cancer_type', mutation_cancer_type)
task.set_parameter('expression_cancer_type', expression_cancer_type)
task.set_parameter('tissue', tissue)
task.set_parameter('include_nutraceutical_drugs', boolean)
task.set_parameter('only_atc_l_drugs', boolean)
task.set_parameter('include_indirect_drugs', boolean)
task.set_parameter('include_non_approved_drugs', boolean)
task.set_parameter('filter_paths', boolean)
task.set_parameter('available_drugs', available_drug_list)
# only avaibale if drug_interaction_dataset_list = ['drugbank']
task.set_parameter('drug_target_action', drug_effect)

List all available datasets

List all available gene interaction datasets:

caddiepy.api.get_gene_interaction_datasets().json()

List all available drug interaction datasets:

caddiepy.api.get_drug_interaction_datasets().json()

List all available tissues:

caddiepy.api.get_tissues().json()

List all available expression cancer types:

caddiepy.api.get_expression_cancer_types().json()

List all available mutation cancer types:

caddiepy.api.get_mutation_cancer_types().json()

List all available drug effects (only relevant when working with DrugBank):

caddiepy.api.get_drug_effects().json()

Look up drugs in the CADDIE database and their interactions with genes:

caddiepy.api.drug_lookup(search_string, database_name)

Logging

Configure the logging level like this:

import logging
logging.basicConfig(level=logging.DEBUG)