Mpro

Generating a SARS-CoV Mpro chemical space with Reinvent¶

Installation

Install Conda
Clone this Git repository: https://github.com/tblaschke/Reinvent-2.0
Open a shell, and go to the repository and create the Conda environment:

In [ ]:

 # conda env create -f reinvent.yml

Activate the environment:

In [ ]:

# conda activate reinvent

Run jupyter notebook

Execute jupyter: jupyter notebook

In [1]:

# load dependencies
import os
import re
import json
import tempfile

# --------- change these path variables as required
reinvent_dir = os.path.expanduser("/mnt/NVMe/Reinvent2-master")
reinvent_env = os.path.expanduser("/home/hazemaj/anaconda3/envs/reinvent_shared.v2.1")
output_dir = os.path.expanduser("/mnt/NVMe/Reinvent2-master/production/SARS_CoV-2_Mpro_5_min_fame")
output_dir_TL = os.path.expanduser("/mnt/NVMe/Reinvent2-master/production/SARS_CoV-2_Mpro_5_min_fame/TL")
output_dir_RL = os.path.expanduser("/mnt/NVMe/Reinvent2-master/production/SARS_CoV-2_Mpro_5_min_fame/RL")
output_dir_sampling = os.path.expanduser("/mnt/NVMe/Reinvent2-master/production/SARS_CoV-2_Mpro_5_min_fame/sampling")

# --------- do not change
# get the notebook's root path
try: ipynb_path
except NameError: ipynb_path = os.getcwd()

# if required, generate a folder to store the results
try:
    os.mkdir(output_dir)
except FileExistsError:
    pass

# if required, generate a folder to store the results
try:
    os.mkdir(output_dir_TL)
except FileExistsError:
    pass

# if required, generate a folder to store the results
try:
    os.mkdir(output_dir_RL)
except FileExistsError:
    pass

# if required, generate a folder to store the results
try:
    os.mkdir(output_dir_sampling)
except FileExistsError:
    pass

In [2]:

#For Transfer Learning: run next cells
#For Reinforcement Learning: Go to Reinforcement Learning and run RL cells
#For Sampling: go to Sampling and run the cells

Transfer Learning¶

In [3]:

# initialize the dictionary
configuration = {
    "version": 2,                          # we are going to use REINVENT's newest release
    "run_type": "transfer_learning"        # other run types: "scoring", "validation",
                                           #                  "transfer_learning",
                                           #                  "reinforcement_learning" and
                                           #                  "create_model"
}

In [4]:

# add block to specify whether to run locally or not and
# where to store the results and logging
configuration["logging"] = {
    "sender": "http://127.0.0.1",          # only relevant if "recipient" is set to "remote"
    "recipient": "local",                  # either to local logging or use a remote REST-interface
    "logging_path": os.path.join(output_dir_TL, "progress.log"), # where the run's output is stored
    "job_name": "Transfer Learning", # set an arbitrary job name for identification
    "job_id": "TL1"                       # only relevant if "recipient" is set to "remote"
}

Define parameters for Transfer Learning
Define the input prior (original prior trained with ChEMBL)
Define the output focused agent path
Define the Mpro chemical space (selection of active molecules from ChEMBL and CovidMoonshot) for which Transfer Leraning should be used
Define the number of epochs

In [5]:

adaptive_lr_config = {
      "mode": "constant", # other modes: "exponential", "adaptive", "constant"
      "gamma": 0.8,
      "step": 1,
      "start": 5E-4,
      "min": 1E-5,
      "threshold": 1E-4,
      "average_steps": 4,
      "patience": 8,
      "restart_value": 1E-5,
      "sample_size": 10000,
      "restart_times": 0
    }

In [6]:

output_model_path = os.path.join(output_dir_TL, "focused.agent") \
# The final focused agent will be named "focused.agent"
# The intermediate steps will be named "focused.agent.1", "focused.agent.2", "focused.agent.3" and etc.

# add the "parameters" block
configuration["parameters"] = {
    "input_model_path": os.path.join(reinvent_dir,        # path to prior or trained agent
                               "data",
                               "augmented.prior"),
    "output_model_path": output_model_path,                # location to store the focused agent
    "input_smiles_path": os.path.join(output_dir,        # path to input smiles                     
                               "SARS-CoV-2_ChemSpace.smi"),# this is a dataset consisting of non-covalent Mpro inhibitors
    "save_every_n_epochs": 1,      # how often to save the focused Agent. Here its stored after each epoch
    "batch_size": 128,             # batch size the input data
    "num_epochs": 20,              # number of epochs to focus the agent for
    "standardize": True,           # the input may contain SMILES strings that are invalid according to the agent
                                   # this atempts to clean up the input dataset
    "randomize": True,             # this triggers data augmentation which is quite important for small datasets
    "adaptive_lr_config": adaptive_lr_config        # setting the learning rate behavior
}

In [7]:

# write the configuration file to the disc
configuration_JSON_path = os.path.join(output_dir_TL, "transfer_learning_config.json")
with open(configuration_JSON_path, 'w') as f:
    json.dump(configuration, f, indent=4, sort_keys=True)

Run REINVENT Now it is time to execute REINVENT locally.

execute REINVENT¶

python /input.py .json

In [8]:

%%capture captured_err_stream --no-stderr

# execute REINVENT from the command-line
!python {reinvent_dir}/input.py {configuration_JSON_path}

Analyse the results¶

In order to analyze the run in a more intuitive way, we can use tensorboard:

go to the root folder of the output cd "your_path"/SARS_CoV-2_Mpro_5_min_fame/TL
make sure, you have activated the proper environment
$conda activate reinvent
start tensorboard
$tensorboard --logdir progress.log

Reinforcement Learning¶

In [9]:

# initialize the dictionary
configuration = {
    "version": 2,                          # we are going to use REINVENT's newest release
    "run_type": "reinforcement_learning"   # other run types: "sampling", "validation",
                                           #                  "transfer_learning",
                                           #                  "scoring" and "create_model"
}

In [10]:

# add block to specify whether to run locally or not and
# where to store the results and logging
configuration["logging"] = {
    "sender": "http://127.0.0.1",          # only relevant if "recipient" is set to "remote"
    "recipient": "local",                  # either to local logging or use a remote REST-interface
    "logging_frequency": 10,               # log every x-th steps
    "logging_path": os.path.join(output_dir_RL, "progress.log"), # load this folder in tensorboard
    "resultdir": os.path.join(output_dir_RL, "results"),         # will hold the compounds (SMILES) and summaries
    "job_name": "Reinforcement Learning",                # set an arbitrary job name for identification
    "job_id": "RL1"                       # only relevant if "recipient" is set to "remote"
}

In [11]:

# add the "parameters" block
configuration["parameters"] = {}

# add a "diversity_filter"
configuration["parameters"]["diversity_filter"] =  {
    "name": "IdenticalMurckoScaffold",     # other options are: "IdenticalTopologicalScaffold", 
                                           #                    "NoFilter" and "ScaffoldSimilarity"
                                           # -> use "NoFilter" to disable this feature
    "nbmax": 25,                           # the bin size; penalization will start once this is exceeded
    "minscore": 0.4,                       # the minimum total score to be considered for binning
    "minsimilarity": 0.4                   # the minimum similarity to be placed into the same bin
}

# prepare the inception (we do not use it in this example, so "smiles" is an empty list)
configuration["parameters"]["inception"] = {
    "smiles": [],                          # fill in a list of SMILES here that can be used (or leave empty)
    "memory_size": 100,                    # sets how many molecules are to be remembered
    "sample_size": 10                      # how many are to be sampled each epoch from the memory
}

# set all "reinforcement learning"-specific run parameters
configuration["parameters"]["reinforcement_learning"] = {
    "prior": os.path.join(reinvent_dir, "data/augmented.prior"), # path to the pre-trained model
    "agent": os.path.join(output_dir_TL, "focused.agent.19"), # path to the pre-trained model
    "n_steps": 500,                        # the number of epochs (steps) to be performed; often 1000
    "sigma": 128,                          # used to calculate the "augmented likelihood", see publication
    "learning_rate": 0.0001,               # sets how strongly the agent is influenced by each epoch
    "batch_size": 128,                     # specifies how many molecules are generated per epoch
    "reset": 0,                            # if not '0', the reset the agent if threshold reached to get
                                           # more diverse solutions
    "reset_score_cutoff": 0.5,             # if resetting is enabled, this is the threshold
    "margin_threshold": 50                 # specify the (positive) margin between agent and prior
}

In [12]:

# prepare the scoring function definition and add at the end
scoring_function = {
    "name": "custom_product",              # this is our default one (alternative: "custom_sum")
    "parallel": True,                     # sets whether components are to be executed
                                           # in parallel; note, that python uses "False" / "True"
                                           # but the JSON "false" / "true"

    # the "parameters" list holds the individual components
    "parameters": [

    # add component: calculate the QED drug-likeness score (using RDkit)
    {
        "component_type": "qed_score",
        "name": "QED Score",                   # arbitrary name for the component
        "weight": 1,                           # the weight of the component (default: 1)
        "model_path": None,                    # not required; note, this is "null" in JSON
        "smiles": [],                          # not required
        "specific_parameters": None            # not required; note, this is "null" in JSON
    },
        
    # add component: enforce the match to a given substructure
    {
        "component_type": "matching_substructure", 
        "name": "Matching substructure",       # arbitrary name for the component
        "weight": 1,                           # the weight of the component (default: 1)
        "model_path": None,                    # not required; note, this is "null" in JSON
        "smiles": ["Brc1ccc2c(c1)CCCO2","CC(=O)N(Cc1ccsc1)c1ccc(N(C)C)cc1","Cc1ccc2nc(N)sc2c1","Cc1ccc2ncsc2c1","CC(C(N)=O)c1cccc(Cl)c1","Cc1ccccc1C1CCC1","Cc1cncc2ccccc12","Cc1nc2n(n1)CC(C(N)=O)CC2","Cc1nc2n(n1)CCCC2","CC(C)(C(N)=O)c1ccccc1","Clc1cc(Cl)c2c(c1)CCCO2","Clc1cc2c(cc1Cl)NCCC2","Clc1cc2c(cc1Cl)OCCC2","Clc1ccc(Cl)cc1","Clc1ccc2c(c1)C=CCC2","Clc1ccc2c(c1)CCCC2","Clc1ccc2c(c1)CCCN2","Clc1ccc2c(c1)CCCO2","Clc1ccc2c(c1)CCCS2","CC(C)(C)NC=O","Clc1ccc2c(c1)CCNC2","Clc1ccc2c(c1)CCO2","Clc1ccc2c(c1Cl)OCCC2","Clc1cccc(CN2CCNCC2)c1","Clc1cccc(Cc2ccn[nH]2)c1","Clc1cccc(Cl)c1","Clc1cccc(N2CCNCC2)c1","Clc1cccc(NCCC2CCCCC2)c1","Clc1cccc(Oc2cccnc2)c1","Clc1cccc2c1OCC2","CC(C)(C)c1ccc(NC(=O)c2cocn2)cc1","Clc1ccccc1Cl","Cn1nc2n(c(=O)c1=O)CCC2","FC(F)c1ccc(Cl)c(Cl)c1","Fc1cc2c(cc1Cl)CCCO2","Fc1ccc(Cl)cc1","Fc1ccc(F)cc1","Fc1ccc2cnccc2c1","Fc1cccc(F)c1","Fc1ccccc1Cl","N#Cc1ccc2cncc(N)c2c1","N#Cc1ccc2cnccc2c1","CC(C)(C)c1cncs1","NC(=O)C1CCOc2ccc(Cl)cc21","NC(=O)CN1CCN(C2CC2)C1=O","NC(=O)Cc1ccc(Cl)c(Cl)c1","NC(=O)Cn1ccnc1","NC(=O)Cn1nc2n(c(=O)c1=O)CCC2","NC(=O)N(CCC1CCCCC1)c1cccc(Cl)c1","NC(=O)[[email protected]@H]1CCOc2ccc(Cl)cc21","NC(=O)[[email protected]]1CCOc2ccc(Cl)cc21","NC(=O)c1cn[nH]c1S(N)(=O)=O","NC1CC(F)(F)C1","NC1CCOCC1","NCC1CCCO1","CC1CCOc2ccc(Cl)cc21","NCc1ccc(Cl)c(Cl)c1","Nc1ccc(C(F)(F)F)cn1","Nc1cncc2c1CCC2","Nc1cncc2c1CNCC2","Nc1cncc2ccc(F)cc12","Nc1cncc2ccccc12","Nc1cncc2ccncc12","CC1Cn2nccc2CN1","Nc1cncc2cnccc12","Nn1nnc2ccccc21","O=C(Cc1ccc(Cl)cc1)Nc1cncc2ccccc12","O=C(Cc1cccc(Cl)c1)Nc1cncc2ccccc12","O=C(Cc1cccc(F)c1)Nc1cncc2ccccc12","O=C(Cc1ccccc1)Nc1cncc2ccccc12","CC1Cn2nccc2CN1C=O","O=C(Cc1cccnc1)NC1CCOC1","O=C(Cc1cccnc1)NC1COC1","O=C(Cc1cncc2ccccc12)NCCC1CCCCC1","O=C(Cn1nnc2ccccc21)NCc1ccsc1","O=C(NCCC1CCCCC1)Nc1cncc2ccccc12","O=C(Nc1ccccc1)C1CC1","O=C(Nc1cccnc1)C1CCOc2ccc(Cl)cc21","O=C(c1ccco1)N1CCNCC1","O=C(c1cncc2ccccc12)N1CCNCC1","O=C1CC(Cc2cccc(Cl)c2)N1","O=C1CC(Oc2cc(O)cc(Cl)c2)N1","O=C1CCCCN1c1cncc2ccccc12","O=C1CCc2cc(Br)ccc2N1","O=C1CCc2cc(Cl)cc(Cl)c2N1","O=C1NCCN1C1CC1","O=CN(CCC1CCCCC1)c1cccc(Cl)c1","O=CN1CCN(Cc2cccc(Cl)c2)CC1","O=CN1CCN(c2cccc(Cl)c2)CC1","O=CNC1CC(F)(F)C1","O=CNC1CCOCC1","O=CNC1COC1","O=CNc1cncc2ccc(F)cc12","O=CNc1cncc2ccccc12","O=CNc1cncc2ccncc12","O=CNc1cncc2cnccc12","O=COc1cncc(Cl)c1","O=c1cccc(O)[nH]1","OCC1CCOc2ccc(Cl)cc21","OCc1ccc(Cl)c(Cl)c1","CCNc1ccc(C#N)cn1","Oc1cncc(Cl)c1","c1cc2c(cn1)CCC2","c1cc2c(cn1)CCNC2","c1cc2ccncc2cn1","c1cc2cnccc2cn1","c1ccc(C2CCCC2)cc1","CCOc1c(C)cccc1Cl","c1ccc2cnccc2c1","c1cnc2ccncc2c1","CCOc1ccccc1Cl","CCS(=O)(=O)N1CCNCC1","CCc1ccc(Cl)c(Cl)c1","CC(=O)N1CCNCC1","CN(C)Cc1ccc(Cl)c(Cl)c1","CN(C)c1ccc(N(C=O)Cc2ccsc2)cc1","CN(C)c1ccc(NC(=O)Cn2nnc3ccccc32)cc1","CN(C)c1ccc(NCc2ccsc2)cc1","CN(C)c1cccnc1","CN1CCCc2cc(Cl)ccc21","CC(=O)N1Cc2ccnn2CC1C","CN1CCN(C2CC2)C1=O","CNC(=O)Cn1ccnc1","CNC(=O)c1cn[nH]c1S(N)(=O)=O","CNCc1ccc(Cl)c(Cl)c1","CNCc1cccc2ccccc12","CNc1ccc(NC(=O)Cn2nnc3ccccc32)cc1","COC(=O)C1(N)CCCC1","COC(=O)C1(NC=O)CCCC1","COC(=O)C1CC(N)C1","COC(=O)C1CC(NC=O)C1","COC(=O)C1CCC1","COC(=O)C1CCCC1","COC(=O)[[email protected]@H]1CC[[email protected]](N)C1","COC(=O)[[email protected]@H]1CC[[email protected]](NC=O)C1","CC(=O)NCc1ccc(Cl)c(Cl)c1","COC1CCOc2ccc(Cl)cc21","COCc1ccc(Cl)c(Cl)c1","COc1cc(Br)ccc1N","COc1cc(Cl)cc(OC2CC(=O)N2)c1","COc1cc(Cl)nc(N)n1","COc1cc(Cl)ncn1","COc1ccc(Br)cn1","COc1ccc(Cl)cc1","COc1ccc(Cl)cc1C","CC(=O)Nc1cncc2ccccc12","COc1ccc2cncc(N)c2c1","COc1ccc2cncc(NC=O)c2c1","COc1ccc2cnccc2c1","COc1cccc(N2CCN(C=O)CC2)c1","COc1cccc(N2CCNCC2)c1","COc1ccccn1","COc1ncc(Br)cc1N","CC(=O)Nc1cncc2ccncc12","CS(=O)(=O)c1ccc2cnccc2c1","Cc1cc(Cl)cc(C#N)c1","Cc1cc(Cl)cc(C(C)C(N)=O)c1","Cc1cc(Cl)cc(Cl)c1","Cc1cc(Cl)cc2c1OCC2","CC(=O)Nc1cncc2cnccc12","Cc1cc(Cl)ccc1Cl","Cc1cc(Cl)ccc1F","Cc1cc(Cl)ccn1","Cc1cc(F)cc(F)c1","Cc1cc(F)ccc1F","Cc1ccc(C)c(Cl)c1","Cc1ccc(Cl)c(Cl)c1","Cc1ccc(Cl)c(F)c1","Cc1ccc(F)c(Cl)c1"],              # a match with this substructure is required
        "specific_parameters": None            # not required; note, this is "null" in JSON
    },
    
    # add component: Get away from the listed structures - SARS-CoV-2 Mpro known inhibitors
    {
        "component_type": "jaccard_distance", 
        "name": "Jaccard distance",         # arbitrary name for the component
        "weight": 1,                           # the weight of the component (default: 1)
        "model_path": None,                    # not required; note, this is "null" in JSON
        "smiles": ['Nc1ccc(O)c2ncccc12', 'Cc1c(-c2cnccn2)ssc1=S', 'O=C(Oc1cncc(Cl)c1)c1ccco1', 'O=C(Oc1cncc(Br)c1)c1ccco1', 'O=C(Oc1cncc(Cl)c1)c1cscn1', 'Cn1sc(=O)n(Cc2ccccc2)c1=O', 'CN1CC[[email protected]@]2(C)c3cc(O)ccc3N(C)[[email protected]@H]12', 'Nc1ccc2cc3ccc(N)cc3nc2c1', 'N#Cc1cnn(C(=O)c2cccs2)c1C1CC1', 'Nc1ccc(S(=O)(=O)[N-]c2ncccn2)cc1', 'COc1cc(C(=S)N2CCOCC2)ccc1O', 'c1ccc2sc(SNC3CCCCC3)nc2c1', 'O=c1oc(Nc2ccccc2)nc2ccccc12', 'Cc1nnn(NC(=O)Cc2cccc(Cl)c2)c1C', 'Nc1ncc(S(=O)(=O)c2ccc(Cl)cc2)c(N)n1', 'COc1cccc(C(=O)Oc2cncc(Cl)c2)c1', 'CCOC(=O)Cc1ccc(-c2ccccc2)cc1', 'O=C(Oc1cncc(Cl)c1)c1cccc2[nH]ccc12', 'COc1cc(Cl)cc(OC(=O)Nc2cccnc2)c1', 'Cc1c(N)cncc1NC(=O)Cc1cccc(Cl)c1', 'O=C(Oc1cncc(Cl)c1)c1cc2ccccc2o1', 'O=C(Oc1cncc(Cl)c1)c1cc2ccccc2[nH]1', 'O=C(Oc1cncc(Cl)c1)c1cc2ccccc2s1', 'C/N=c1/nc(-c2ccccc2)n(-c2ccccc2)s1', 'CC(CN1CC(=O)NC(=O)C1)N1CC(=O)NC(=O)C1', 'Cc1ccc(C)c(-n2sc3cc(F)ccc3c2=O)c1', 'O=C(Cc1cccc(Cl)c1)Nc1cncc2c1CCC2', 'O=C(O)[[email protected]@H](c1ccccc1)N1Cc2ccccc2C1=O', 'N#Cc1ccc(CN2CCC(C3CCNC3)CC2)cc1', 'O=C(Cc1cccc(Cl)c1)Nc1cncc2nc[nH]c12', 'O=C(Cc1cccc(Cl)c1)Nn1cnc2ccccc21', 'O=C(Cc1cccc(Cl)c1)Nc1cnccc1C1CC1', 'O=C(Cc1ccccc1)Nc1cncc2ccccc12', 'O=C(Cc1cccc(Cl)c1)Nn1nnc2ccccc21', 'O=C(Cc1cccc(Cl)c1)Nc1cnc2ccccn12', 'CCCCc1nc(-c2nccn2Cc2ccoc2)c[nH]1', 'CC(=O)N1CCN(CC(=O)Nc2cnccc2C)CC1', 'O=C(CC1Sc2ncnn2C1=O)Nc1ccc(Br)cc1', 'O=C(O)C(c1ccccc1)N1Cc2ccccc2C1=O', 'O=C(Cc1cncc2ccccc12)Nc1cccc(Cl)c1', 'O=C(Cc1cc(Cl)ccn1)Nc1cncc2ccccc12', 'O=C(Cc1cccc(Cl)c1)Nc1cncc2c1CCCN2', 'Cc1ccncc1NC(=O)[[email protected]]1CCOc2ccc(Cl)cc21', 'O=C(Cc1cccc(Cl)c1)Nc1cncc2ccncc12', 'O=C(Cc1cccc(Cl)c1)Nc1cncc2cnccc12', 'O=C(Cc1cccc(Cl)c1)Nc1cncc2cccnc12', 'O=C(Cc1cccc(Cl)c1)Nc1cncc2c1OCCC2', 'O=C(Cc1cccc(Cl)c1)Nc1cncc2ccccc12', 'O=C(Cc1ccc(F)cc1)Nc1cncc2ccccc12', 'O=C(Cc1cccc(F)c1)Nc1cncc2ccccc12', 'Cc1cccc(CC(=O)Nc2cncc3ccccc23)c1', 'O=C(Cc1cncc(Cl)c1)Nc1cncc2ccccc12', 'O=C(Cc1ccc(Cl)cn1)Nc1cncc2ccccc12', 'Cc1ccncc1NC(=O)C1CCOc2ccc(Cl)cc21', 'Cc1cc(C)cc(N2C(=O)CC(Sc3nc[nH]n3)C2=O)c1', 'O=C(Cc1ccc(Cl)c(Cl)c1)Nc1cncc2ccccc12', 'O=C(COc1cccc(Cl)c1)Nc1cncc2ccccc12', 'Cc1ccncc1NC(=O)Cc1cc(Cl)cc(S(N)(=O)=O)c1', 'O=C(Cc1ccc(Cl)c(Cl)c1)Nc1cncc2cnccc12', 'O=C(Cc1ccc(Cl)c(Cl)c1)Nc1cncc2ccncc12', 'O=C(Cc1cc(Cl)cc(Cl)c1)Nc1cncc2ccccc12', 'O=C(Cc1ccc(F)c(Cl)c1)Nc1cncc2ccccc12', 'Cc1ccc(CC(=O)Nc2cncc3ccccc23)cc1Cl', 'O=C(Cc1ccc(Cl)c(F)c1)Nc1cncc2ccccc12', 'N#Cc1cccc(CC(=O)Nc2cncc3ccccc23)c1', 'O=C(Cc1cc(F)cc(Cl)c1)Nc1cncc2ccccc12', 'O=C(Cc1cc(F)cc(F)c1)Nc1cncc2ccccc12', 'O=C(Cc1cccc(F)c1F)Nc1cncc2ccccc12', 'O=C(Cc1cc(F)ccc1F)Nc1cncc2ccccc12', 'O=C(Cc1cc(Cl)ccc1F)Nc1cncc2ccccc12', 'O=C(Cc1cc(Cl)ccc1Cl)Nc1cncc2ccccc12', 'C#Cc1cccc(CC(=O)Nc2cncc3ccccc23)c1', 'O=C(Nc1cnccc1CO)C1CCOc2ccc(Cl)cc21', 'O=C(Nc1nncn1C1CC1)C1CCOc2ccc(Cl)cc21', 'O=C(Cc1cccc(Cl)c1)Nn1cnc2ccccc2c1=O', 'O=C(Oc1cncc(Cl)c1)c1ccc(-c2ccc(Cl)cc2)o1', 'Cc1ccc(-n2sc(=O)n(Cc3ccc(F)cc3)c2=O)cc1', 'CC(C)c1ccc2oc3nc(N)c(C(=O)O)cc3c(=O)c2c1', 'O=C(Nc1cncc2ccccc12)C1COc2c(Cl)cccc21', 'COC(=O)c1c(C)nn(C(=O)c2cccs2)c1-c1snnc1C', 'O=C(Nc1cncc2ccccc12)C1COc2ccc(Cl)cc21', 'O=C1C(c2cccc(Cl)c2)CCN1c1cncc2ccccc12', 'COc1ccc(Cl)cc1CC(=O)Nc1cncc2ccccc12', 'COc1ccc2cncc(NC(=O)Cc3cccc(Cl)c3)c2c1', 'O=C1N(Cc2cccc(Cl)c2)CCN1c1nccc2[nH]ncc12', 'NC(C(=O)Nc1cncc2ccccc12)c1ccc(Cl)c(Cl)c1', 'CC(C(=O)Nc1cncc2ccccc12)c1ccc(Cl)c(Cl)c1', 'O=C(Nc1cncc2ccccc12)C(O)c1ccc(Cl)c(Cl)c1', 'N#Cc1cc(Cl)cc(CC(=O)Nc2cncc3ccccc23)c1', 'O=C(Nc1cncc2ccccc12)[[email protected]@H]1COc2ccc(Cl)cc21', 'O=[N+]([O-])c1cc(C(F)(F)F)ccc1S(=O)(=O)c1ccc(Cl)cc1', 'OC1OC(c2cccs2)=N/C1=C/c1ccc(N2CCOCC2)s1', 'O=C(Cc1cncc2ccccc12)N1CCOc2ccc(Cl)cc21', 'O=C(Nc1cncc2ccccc12)C1=CCCc2ccc(Cl)cc21', 'COc1ccccc1OCCNC(=O)c1cncc2ccccc12', 'O=C(c1cncc2ccccc12)N1CCN(c2ccccc2)CC1', 'O=C(Nc1cncc2ccccc12)C1CCCc2ccc(Cl)cc21', 'O=C(Nc1cncc2ccccc12)C1CCNc2ccc(Cl)cc21', 'O=C(Nc1cncc2ccccc12)C1CCOc2ccc(Br)cc21', 'O=C(Cc1cc(Cl)cc2c1OCC2)Nc1cncc2ccccc12', 'O=C(Nc1cncc2ccccc12)[[email protected]@H]1CCOc2ccc(Cl)cc21', 'O=C(Nc1cncc2ccccc12)C1CCOc2ccc(Cl)cc21', 'O=C1C(c2cccc(Cl)c2)CCCN1c1cncc2ccccc12', 'Cc1ccncc1NC(=O)Cc1cc(Cl)cc(OC2CC(=O)N2)c1', 'O=C(Nc1cncc2ccccc12)C1CCSc2ccc(Cl)cc21', 'COC(C(=O)Nc1cncc2ccccc12)c1ccc(Cl)c(Cl)c1', 'CCOc1c(Cl)cccc1CC(=O)Nc1cncc2ccccc12', 'CNC(C(=O)Nc1cncc2ccccc12)c1ccc(Cl)c(Cl)c1', 'O=C(Nc1cncc2ccccc12)C(F)(F)c1ccc(Cl)c(Cl)c1', 'Cn1cnc2cncc(NC(=O)C3CCOc4ccc(Cl)cc43)c21', 'O=C(Cc1cccc(C(F)(F)F)c1)Nc1cncc2ccccc12', 'O=C(Nc1cncc2ccccc12)[[email protected]@H]1CCNc2ccc(Cl)cc21', 'O=C(Nc1cncc2ccncc12)[[email protected]]1CCOc2ccc(Cl)cc21', 'O=C(Nc1cncc2c1CNCC2)C1CCOc2ccc(Cl)cc21', 'O=C(Nc1cncc2ccccc12)N1CCOc2ccc(Cl)cc21', 'O=C(Nc1cncc2ccccc12)C1CNCc2ccc(Cl)cc21', 'O=C(Nc1cncc2cnccc12)[[email protected]@H]1CCOc2ccc(Cl)cc21', 'O=C(Nc1cncc2ccncc12)[[email protected]@H]1CCOc2ccc(Cl)cc21', 'OCCn1c(CSc2nc3ccccc3o2)nc2ccc(Cl)cc21', 'Cn1nc(-c2ccc(C(F)(F)F)cc2)nc2c(=O)n(C)c(=O)nc1-2', 'CSc1[nH]nc(NC(=O)c2cccs2)c1S(=O)(=O)c1ccccc1', 'O=c1sn(-c2cccc3ccccc23)c(=O)n1Cc1ccccc1', 'O=C(Nc1cncc2cnccc12)C1CCOc2ccc(Cl)cc21', 'O=C(Nc1cncc2ccncc12)C1CCOc2ccc(Cl)cc21', 'O=C(Nc1cncc2ccc(F)cc12)C1CCOc2ccc(Cl)cc21', 'O=C(c1cncc2ccccc12)N1CCN(c2cccc(Cl)c2)CC1', 'O=C1CC(C(=O)Nc2cncc3ccccc23)c2cc(Br)ccc2N1', 'C[[email protected]@H]1C[[email protected]](C(=O)Nc2cncc3ccccc23)c2cc(Cl)ccc2O1', 'C[[email protected]@H]1COc2ccc(Cl)cc2[[email protected]@H]1C(=O)Nc1cncc2ccccc12', 'C[[email protected]@H]1C[[email protected]@H](C(=O)Nc2cncc3ccccc23)c2cc(Cl)ccc2O1', 'C[[email protected]]1COc2ccc(Cl)cc2[[email protected]@H]1C(=O)Nc1cncc2ccccc12', 'C[[email protected]@]1(C(=O)Nc2cncc3ccccc23)CCOc2ccc(Cl)cc21', 'Cc1ccncc1NC(=O)Cc1cc(C#N)cc(OC2CC(=O)N2)c1', 'COc1ccccc1OCCNC(=O)c1cc(=O)[nH]c2ccccc12', 'O=C(Nc1cccnc1)N(CCC1CCCCC1)c1cccc(Cl)c1', 'O=C(Nc1cncc2ccccc12)[[email protected]@H]1CCOc2cc(F)c(Cl)cc21', 'N[[email protected]]1(C(=O)Nc2cncc3ccccc23)CCOc2ccc(Cl)cc21', 'O=C(Nc1cncc2ccccc12)[[email protected]@H]1CCNc2cc(Cl)c(Cl)cc21', 'O=C(Nc1cncc2ccc(F)cc12)[[email protected]@H]1CCOc2ccc(Cl)cc21', 'CN1CC[[email protected]](C(=O)Nc2cncc3ccccc23)c2cc(Cl)ccc21', 'CN(C)C(C(=O)Nc1cncc2ccccc12)c1ccc(Cl)c(Cl)c1', 'O=C(Nc1cncc2ccccc12)C1CCOc2cc(F)c(Cl)cc21', 'O=C1C(c2ccc(Cl)c(Cl)c2)CCCN1c1cncc2ccccc12', 'O=C(Nc1cncc2ccccc12)C1CCOc2c(Cl)cc(Cl)cc21', 'O=C(Nc1cncc2ccccc12)C1CCNc2cc(Cl)c(Cl)cc21', 'O=C(Nc1cncc2ccccc12)C1CCOc2cc(Cl)c(Cl)cc21', 'O=C(Nc1cncc2ccccc12)C1CCOc2c1ccc(Cl)c2Cl', 'O=C(Nc1cncc2ccccc12)[[email protected]]1(O)CCOc2ccc(Cl)cc21', 'O=C(Nc1cncc2ccccc12)C1CCOc2c(O)cc(Cl)cc21', 'O=C1CC(C(=O)Nc2cncc3ccccc23)c2cc(Cl)ccc2N1', 'CN1CCC(C(=O)Nc2cncc3ccccc23)c2cc(Cl)ccc21', 'Cc1c(Cl)ccc2c1[[email protected]](C(=O)Nc1cncc3ccccc13)CCO2', 'CCn1c(-c2nonc2N)nc2cncc(CNC3CCNCC3)c21', 'CC1(C(=O)Nc2cncc3ccccc23)CCOc2ccc(Cl)cc21', 'O=C(c1cncc2ccccc12)N1CCN(c2cccc(Cl)c2)C(=O)C1', 'O=C(c1cc(=O)[nH]c2ccccc12)N1CCN(c2cccc(Cl)c2)CC1', 'c1ccc(-c2cc(Sc3nnc(-c4cccs4)o3)n3ncnc3n2)cc1', 'COc1ccccc1OCCNC(=O)c1cc(=O)[nH]c2cccc(C)c12', 'O=C(Nc1cncc2ccccc12)C1(CO)CCOc2ccc(Cl)cc21', 'O=C1CC(C(=O)Nc2cncc3ccccc23)c2cc(Cl)cc(Cl)c2N1', 'COc1ccc2cncc(NC(=O)C3CCOc4ccc(Cl)cc43)c2c1', 'COC1(C(=O)Nc2cncc3ccccc23)CCOc2ccc(Cl)cc21', 'CC(=O)NC(C(=O)Nc1cncc2ccccc12)c1ccc(Cl)c(Cl)c1', 'CC(C(N)=O)c1cc(Cl)cc(CC(=O)Nc2cncc3ccccc23)c1', 'NC[[email protected]@]1(C(=O)Nc2cncc3ccccc23)CCOc2ccc(Cl)cc21', 'N#Cc1ccc2cncc(NC(=O)C3CCOc4ccc(Cl)cc43)c2c1', 'CO[[email protected]@]1(C(=O)Nc2cncc3ccccc23)CCOc2ccc(Cl)cc21', 'O=C(Cc1cc(Cl)cc(-c2nnc[nH]2)c1)Nc1cncc2ccccc12', 'COc1c(Cl)cccc1OCCNC(=O)c1cc(=O)[nH]c2ccccc12', 'O=C(c1cc(=O)[nH]c2ccccc12)N1CCN(Cc2ccccc2)CC1', 'C[[email protected]]1COc2c(N3CCN(C)CC3)c(F)cc3c(=O)c(C(=O)O)cn1c23', 'O=C(c1cc(O)nc2ccccc12)N1CCN(c2cccc(Cl)c2)CC1', 'Cc1ccc(N(Cc2cscn2)C(=O)Cc2cncc3ccccc23)cc1', 'O=C(Cc1cc(Cl)cc(CC2CC(=O)N2)c1)Nc1cncc2ccccc12', 'O=C(Cc1cc(Cl)cc(Cc2ccn[nH]2)c1)Nc1cncc2ccccc12', 'Cc1cc(C)c(C#N)c(Sc2ncnc3c2nnn3Cc2ccccc2)n1', 'O=C(Cc1cc(Cl)cc(OC2CC(=O)N2)c1)Nc1cncc2ccccc12', 'CO[[email protected]@]1(C(=O)Nc2cncc3ccccc23)CCOc2c(F)cc(F)cc21', 'COC1(C(=O)Nc2cncc3ccccc23)CCOc2c(F)cc(F)cc21', 'COC1(C(=O)Nc2cncc3ccccc23)CCOc2cc(F)c(Cl)cc21', 'CC(=O)N1CC[[email protected]](C(=O)Nc2cncc3ccccc23)c2cc(Cl)ccc21', 'CO[[email protected]]1(C(=O)Nc2cncc3ccccc23)CCOc2cc(Cl)c(Cl)cc21', 'COC1(C(=O)Nc2cncc3ccccc23)CCOc2cc(Cl)c(Cl)cc21', 'O=C(Nc1cncc2ccccc12)C(NCC1CC1)c1ccc(Cl)c(Cl)c1', 'COCCOC(C(=O)Nc1cncc2ccccc12)c1ccc(Cl)c(Cl)c1', 'NC(=O)N1CCC(C(=O)Nc2cncc3ccccc23)c2cc(Cl)ccc21', 'CC1CN(C(=O)c2cc(=O)[nH]c3ccccc23)CCN1c1cccc(Cl)c1', 'CC(=O)N1CCC(C(=O)Nc2cncc3ccccc23)c2cc(Cl)ccc21', 'O=C1CC2(CCOc3ccc(Cl)cc32)C(=O)N1c1cncc2ccccc12', 'O=C(c1cc(=O)[nH]c2ccccc12)N1CCN(Cc2cccc(Cl)c2)CC1', 'CNc1ccc(N(Cc2ccsc2)C(=O)Cn2nnc3ccccc32)cc1', 'COc1cccc(N2CCN(C(=O)c3cc(=O)[nH]c4ccccc34)CC2)c1', 'Cc1noc(NC(=O)c2ccc(-c3cc(C(F)(F)F)nn3C)s2)c1[N+](=O)[O-]', 'COc1cccc(N2CCN(C(=O)c3cc(O)nc4ccccc34)CC2)c1', 'O=C(c1cc(O)nc2ccccc12)N1CCN(Cc2cccc(Cl)c2)CC1', 'COc1cccc2[nH]c(=O)cc(C(=O)N3CCN(c4cccc(Cl)c4)CC3)c12', 'Cc1ccc(N(Cc2cccc(Cl)c2)C(=O)Cn2nnc3ccccc32)cc1', 'O=C(Cc1cncc2ccccc12)N(CCC1CCCCC1)Cc1cccs1', 'CN(C)c1ccc(N(Cc2cscn2)C(=O)Cn2nnc3ccccc32)cc1', 'O=C(Cc1cc(Cl)cc(Oc2cccnc2)c1)Nc1cncc2ccccc12', 'CNc1ccc(N(C(=O)Cn2nnc3ccccc32)c2cccc(Cl)c2)cc1', 'CN(C)c1ccc(N(Cc2ccsc2)C(=O)Cn2nnc3ccccc32)cc1', 'COC[[email protected]]1(C(=O)Nc2cncc3ccccc23)CCOc2cc(Cl)c(Cl)cc21', 'COCC1(C(=O)Nc2cncc3ccccc23)CCOc2cc(Cl)c(Cl)cc21', 'CC(NC(=O)CN)(C(=O)Nc1cncc2ccccc12)c1ccc(Cl)c(Cl)c1', 'COCCN1CCC(C(=O)Nc2cncc3ccccc23)c2cc(Cl)ccc21', 'CS(=O)(=O)c1ccc2cncc(NC(=O)C3CCOc4ccc(Cl)cc43)c2c1', 'Cc1ccncc1NC(=O)Cc1cc(Cl)cc(NC2(C)CCC(F)(F)CC2)c1', 'CN(C)c1ccc(N(Cc2ccsc2)C(=O)CN2N=NC3C=CC=CC32)cc1', 'CCC(=O)N(C(=O)C1CCOc2ccc(Cl)cc21)c1cncc2ccccc12', 'COc1ccc(N(Cc2cccc(Cl)c2)C(=O)Cn2nnc3ccccc32)cc1', 'O=C(c1cc(=O)[nH]c2ccccc12)N1CCN(c2cccc(C(F)(F)F)c2)CC1', 'O=C(Cc1cncc2ccccc12)N(CCC1CCCCC1)c1cccc(Cl)c1', 'CN(C)c1ccc(N(Cc2ccsc2)C(=O)Cc2cncc3ccccc23)cc1', 'CN(C)c1ccc(N(Cc2ccccc2)C(=O)Cn2nnc3ccccc32)cc1', 'Cc1ccncc1NC(=O)Cc1cc(Cl)cc(-c2ccc(C3CC3(F)F)cc2)c1', 'O=C(Nc1cncc2ccccc12)N(CCC1CCCCC1)c1cccc(Cl)c1', 'CNC(=O)C[[email protected]]1(C(=O)Nc2cncc3ccccc23)CCOc2ccc(Cl)cc21', 'COCC(=O)NC(C)(C(=O)Nc1cncc2ccccc12)c1ccc(Cl)c(Cl)c1', 'CNC(=O)CC1(C(=O)Nc2cncc3ccccc23)CCOc2ccc(Cl)cc21', 'Cc1ccncc1NC(=O)Cc1cc(Cl)cc(-c2ccc(C3(C)CCO3)cc2)c1', 'O=C(Cc1cc(Cl)cc(Oc2cccc(=O)[nH]2)c1)Nc1cncc2ccccc12', 'CN(C)S(=O)(=O)c1cccc(C(=O)Nc2c(-c3ccccc3)nc3sccn23)c1', 'COC(=O)c1c(C)n(S(C)(=O)=O)c2ccc(N(C(=O)c3ccco3)S(C)(=O)=O)cc12', 'CN(C)c1ccc(N(Cc2cccc(Cl)c2)C(=O)Cn2nnc3ccccc32)cc1', 'COC(=O)c1ccc(-c2cc(Cl)cc(CC(=O)Nc3cnccc3C)c2)c(OC)c1', 'COc1cccc(Oc2cc(Cl)cc(NC(=O)Cc3cncc4ccccc34)c2)n1', 'O=C1CC(Oc2cc(Cl)cc(OCCNC(=O)c3cc(=O)[nH]c4ccccc34)c2)N1', 'CCC(=O)Nc1ccc(N(Cc2ccsc2)C(=O)Cn2nnc3ccccc32)cc1', 'O=C(Cc1cc(Cl)cc(NCc2ccc(Br)cc2)c1)Nc1cncc2ccccc12', 'CN(C)C(=O)C[[email protected]]1(C(=O)Nc2cncc3ccccc23)CCOc2ccc(Cl)cc21', 'CN(C)C(=O)CC1(C(=O)Nc2cncc3ccccc23)CCOc2ccc(Cl)cc21', 'O=C(Nc1cncc2ccccc12)C1CCN(Cc2ncc[nH]2)c2ccc(Cl)cc21', 'O=C(Nc1cncc2ccccc12)C1CCN(Cc2ccn[nH]2)c2ccc(Cl)cc21', 'CN(C)C(=O)CN1CCC(C(=O)Nc2cncc3ccccc23)c2cc(Cl)ccc21', 'O=C(Nc1cncc2ccccc12)C1CCN(Cc2ncn[nH]2)c2ccc(Cl)cc21', 'O=C(Nc1cncc2ccccc12)C1CCN(Cc2cnc[nH]2)c2ccc(Cl)cc21', 'CNC(=O)Nc1ccc(N(Cc2ccsc2)C(=O)Cn2nnc3ccccc32)cc1', 'CCOC(=O)c1c(C)n(S(C)(=O)=O)c2ccc(N(C(=O)c3ccco3)S(C)(=O)=O)cc12', 'CCOC(=O)c1cc(-c2ccccc2)sc1NC(=O)Cc1csc(-c2ccccc2)n1', 'CCC(=O)Nc1ccc(N(Cc2ccccc2)C(=O)Nn2nnc3ccccc32)cc1', 'O=C(Cn1nnc2ccccc21)N(Cc1ccsc1)c1ccc(N2CCCCC2)cc1', 'COc1cc(Cl)nc(Nc2cc(Cl)cc(CC(=O)Nc3cncc4ccccc34)c2)n1', 'COc1cc(Br)ccc1Nc1cc(Cl)cc(CC(=O)Nc2cncc3ccccc23)c1', 'COc1ncc(Br)cc1Nc1cc(Cl)cc(CC(=O)Nc2cncc3ccccc23)c1', 'CN(C)c1ccc(N(Cc2cccc(Cl)c2)C(=O)Cc2cncc3ccccc23)nc1', 'O=C(Nc1cncc2ccccc12)C1CCN(Cc2ccccc2)c2ccc(Cl)cc21', 'O=C(Cn1nnc2ccccc21)N(Cc1ccsc1)c1ccc(N2CCOCC2)cc1', 'COc1cccc(CCN(C(=O)c2cc(=O)[nH]c3ccccc23)c2cccc(Cl)c2)c1', 'O=C(Cn1nnc2ccccc21)N(Cc1ccsc1)c1ccc(-c2ccccc2)cc1', 'O=C(Nc1ccc(N(Cc2ccsc2)C(=O)Cn2nnc3ccccc32)cc1)C1CC1', 'CC(C)(C(N)=O)c1ccc(N(Cc2ccsc2)C(=O)Cn2nnc3ccccc32)cc1', 'O=S(=O)(Cc1ccccc1F)c1oc(-c2ccc(F)cc2)nc1S(=O)(=O)c1ccccc1', 'Cc1cc(C)c(C)c(S(=O)(=O)N2CCN(C(=O)c3cc(=O)[nH]c4ccccc34)CC2)c1C', 'O=C(Cc1cc(Cl)cc(Nc2ccc(C(F)(F)F)cn2)c1)Nc1cncc2ccccc12', 'CCN(Cc1ccc(F)cc1)c1cc(Cl)cc(CC(=O)Nc2cncc3ccccc23)c1', 'Cc1ccc2nc(Nc3cc(Cl)cc(CC(=O)Nc4cncc5ccccc45)c3)sc2c1', 'CCS(=O)(=O)N1CCN(c2cc(Cl)cc(CC(=O)Nc3cncc4ccccc34)c2)CC1', 'CN(C)CC(=O)NCC1(C(=O)Nc2cncc3ccccc23)CCOc2ccc(Cl)cc21', 'COCCNC(=O)C(c1cccnc1)N(C(=O)c1cocn1)c1ccc(C(C)(C)C)cc1', 'CC(C)(C)c1ccc(N(C(=O)c2cocn2)C(C(=O)NC2COC2)c2cccnc2)cc1', 'CC(=O)Nc1ccc(N(Cc2cccc(Cl)c2)C(=O)Cc2cncc3ccccc23)cn1', 'CC(C)(C)NC(=O)C(c1cccnc1)N(C(=O)c1ccco1)c1ccc(C(C)(C)C)cc1', 'CC(C)c1cc(CNc2nc(Nc3cc(C4CC4)[nH]n3)cc(N3CCN(C)CC3)n2)on1', 'CC[[email protected]](CO)Nc1nc(NCc2ccc(-c3ccccn3)cc2)c2ncn(C(C)C)c2n1', 'CC(C)Nc1cccnc1N1CCN(C(=O)c2cc3cc(NS(C)(=O)=O)ccc3[nH]2)CC1', 'O=C(Cn1nnc2ccccc21)N(Cc1cccc(Cl)c1)c1ccc(N2CCOCC2)cc1', 'O=C(Cn1nnc2ccccc21)N(Cc1cccc(Cl)c1)c1ccc(N2CCCCC2)cc1', 'CC(C)(C)c1ccc(N(C(=O)c2cocn2)C(C(=O)NC2CCOC2)c2cccnc2)cc1', 'COC(=O)CNC(=O)C(c1cccnc1)N(C(=O)c1cocn1)c1ccc(C(C)(C)C)cc1', 'Cn1ccc(CNCC2(C(=O)Nc3cncc4ccccc34)CCOc3ccc(Cl)cc32)n1', 'Cc1nc2n(n1)CC(C(=O)NC(C(=O)Nc1cncc3ccccc13)c1cccc(Cl)c1)CC2', 'O=C(Cc1cc(Cl)cc(N2CCN(C(=O)c3ccco3)CC2)c1)Nc1cncc2ccccc12', 'CN(Cc1cccc2ccccc12)c1cc(Cl)cc(CC(=O)Nc2cncc3ccccc23)c1', 'O=C(Cn1ccnc1)NCC1(C(=O)Nc2cncc3ccccc23)CCOc2ccc(Cl)cc21', 'COC(=O)C(C)NC(=O)C(c1cccnc1)N(C(=O)c1cocn1)c1ccc(C(C)(C)C)cc1', 'CC(C)(C)c1ccc(N(C(=O)c2cocn2)C(C(=O)NC2CCOCC2)c2cccnc2)cc1', 'CC(C)(C)c1ccc(N(C(=O)c2cocn2)C(C(=O)NCC2CCOC2)c2cccnc2)cc1', 'CC(C)(C)c1ccc(N(C(=O)c2cocn2)C(C(=O)NCC2CCCO2)c2cccnc2)cc1', 'CC(C)(C)c1ccc(N(C(=O)c2cocn2)C(C(=O)NC2CC(F)(F)C2)c2cccnc2)cc1', 'Cn1c(=O)ccn(CC(=O)NC(C(=O)Nc2cncc3ccccc23)c2cccc(Cl)c2)c1=O', 'O=C(Cn1cnc(C(F)(F)F)n1)NC(C(=O)Nc1cncc2ccccc12)c1cccc(Cl)c1', 'O=C(Cn1cc(Cl)c(=O)[nH]c1=O)NC(C(=O)Nc1cncc2ccccc12)c1cccc(Cl)c1', 'O=C(CN1CCN(C2CC2)C1=O)NC(C(=O)Nc1cncc2ccccc12)c1cccc(Cl)c1', 'Cn1nc(C(=O)NC(C)(C(=O)Nc2cncc3ccccc23)c2ccc(Cl)c(Cl)c2)cc1C#N', 'CC(NC(=O)c1cc2n(n1)CCO2)(C(=O)Nc1cncc2ccccc12)c1ccc(Cl)c(Cl)c1', 'CC(C)(C)c1ccc(N(C(=O)c2ccco2)C(C(=O)Nc2ccccc2)c2cccnc2)cc1', 'O=C1CC(Oc2cc(Cl)cc(N(CCC3CCCCC3)C(=O)Nc3cncc4ccccc34)c2)N1', 'Cn1cc(C2=NOC(C(=O)NC(C(=O)Nc3cncc4ccccc34)c3cccc(Cl)c3)C2)cn1', 'CC1CN(C)CCN1C(=O)CC1(C(=O)Nc2cncc3ccccc23)CCOc2ccc(Cl)cc21', 'CC(C)N1CCOC(C(=O)NC(C)(C(=O)Nc2cncc3ccccc23)c2ccc(Cl)c(Cl)c2)C1', 'CC(C)(C)c1ccc(N(C(=O)c2ccco2)C(C(=O)NCc2ccccc2)c2cccnc2)cc1', 'N#Cc1ccc(N(C(=O)c2ccco2)C(C(=O)NCCc2cccc(F)c2)c2cccnc2)cc1', 'COC(=O)C1CC(NC(=O)C(c2cccnc2)N(C(=O)c2cocn2)c2ccc(C(C)(C)C)cc2)C1', 'COc1ccc(NC(=O)C(c2cccnc2)N(C(=O)c2cocn2)c2ccc(C(C)(C)C)cc2)cn1', 'O=C(Cn1nc2n(c(=O)c1=O)CCC2)NC(C(=O)Nc1cncc2ccccc12)c1cccc(Cl)c1', 'O=C(Cc1ccc(-n2cnnn2)cc1)NC(C(=O)Nc1cncc2ccccc12)c1cccc(Cl)c1', 'CC(NC(=O)CN1CCN(C2CC2)C1=O)(C(=O)Nc1cncc2ccccc12)c1ccc(Cl)c(Cl)c1', 'Cc1nc2n(n1)CC(C(=O)NC(C)(C(=O)Nc1cncc3ccccc13)c1ccc(Cl)c(Cl)c1)CC2', 'CC(C)(C)c1ccc(N(C(=O)c2ccco2)C(C(=O)NCCN2CCOCC2)c2cccnc2)cc1', 'CC(C)(C)c1cnc(N(C(=O)c2ccco2)C(C(=O)NCCc2cccc(F)c2)c2cccnc2)s1', 'CC(C)(C)c1ccc(N(C(=O)c2ccco2)[[email protected]](C(=O)NCCc2cccc(F)c2)c2cccnc2)cc1', 'CC(C)(C)c1ccc(N(C(=O)c2c[nH]cn2)C(C(=O)NCCc2cccc(F)c2)c2cccnc2)cc1', 'CC(C)(C)c1ccc(N(C(=O)c2cn[nH]c2)C(C(=O)NCCc2cccc(F)c2)c2cccnc2)cc1', 'CC(C)(C)c1ccc(N(C(=O)c2ccco2)C(C(=O)NCCc2cccc(F)c2)c2cccnc2)cc1', 'CC(C)(C)c1ccc(N(C(=O)c2cocn2)C(C(=O)NCCc2cccc(F)c2)c2ccnnc2)cc1', 'CC(C)(C)c1ccc(N(C(=O)c2cocn2)C(C(=O)NCCc2cccc(F)c2)c2cncnc2)cc1', 'CC(C)(C)c1ccc(N(C(=O)c2cocn2)C(C(=O)NCCc2cccc(F)c2)c2cnccn2)cc1', 'O=C(CC1(C(=O)Nc2cncc3ccccc23)CCOc2ccc(Cl)cc21)N1CCN(CCCO)CC1', 'COC(=O)[[email protected]@H]1CC[[email protected]](NC(=O)C(c2cccnc2)N(C(=O)c2cocn2)c2ccc(C(C)(C)C)cc2)C1', 'COC(=O)C1(NC(=O)C(c2cccnc2)N(C(=O)c2cocn2)c2ccc(C(C)(C)C)cc2)CCCC1', 'CC1Cn2nccc2CN1C(=O)CC1(C(=O)Nc2cncc3ccccc23)CCOc2ccc(Cl)cc21', 'CC(C)Oc1ccc(N(C(=O)c2cocn2)C(C(=O)NCCc2cccc(F)c2)c2cccnc2)cc1', 'CC(C)(C)c1ccc(N(C(=O)c2c[nH]nn2)C(C(=O)NCCc2cccc(F)c2)c2cccnc2)cc1', 'CC(C)(C)c1ccc(N(C(=O)c2cnco2)C(C(=O)NCCc2cccc(F)c2)c2cccnc2)cc1', 'NS(=O)(=O)c1[nH]ncc1C(=O)NC[[email protected]@]1(C(=O)Nc2cncc3ccccc23)CCOc2ccc(Cl)cc21', 'CC(C)Oc1ccc(N(C(=O)c2c[nH]cn2)C(C(=O)NCCc2cccc(F)c2)c2cccnc2)cc1', 'CC(C)(C)c1ccc(N(C(=O)C2CCCO2)C(C(=O)NCCc2cccc(F)c2)c2cccnc2)cc1', 'CC(C)(C)c1ccc(N(C(=O)c2ncc[nH]2)C(C(=O)NCCc2cccc(F)c2)c2cccnc2)cc1', 'CC(C)(C)c1ccc(N(C(=O)c2nccs2)C(C(=O)NCCc2cccc(F)c2)c2cccnc2)cc1', 'CC(C)(C)c1ccc(N(C(=O)c2cocn2)C(C(=O)NCCc2cccc(F)c2)c2cccnc2)cc1', 'CC(C)(C)c1ccc(N(C(=O)c2ccn[nH]2)C(C(=O)NCCc2cccc(F)c2)c2cccnc2)cc1', 'COc1nc2ccc(Br)cc2cc1[[email protected]@H](c1ccccc1)[[email protected]@](O)(CCN(C)C)c1cccc2ccccc12', 'CC(C)(C)c1ccc(N(C(=O)c2ccc(Cl)o2)C(C(=O)NCCc2cccc(F)c2)c2cccnc2)cc1', 'Cc1ccncc1C(C(=O)NCCc1cccc(F)c1)N(C(=O)c1cocn1)c1ccc(C(C)(C)C)cc1', 'O=C(CC1(C(=O)Nc2cncc3ccccc23)CCOc2ccc(Cl)cc21)N1CCCCC1c1cn[nH]c1', 'O=C(CC1(C(=O)Nc2cncc3ccccc23)CCOc2ccc(Cl)cc21)N1CCN(C2CCOC2)CC1', 'O=C(C[[email protected]]1(C(=O)Nc2cncc3ccccc23)CCOc2ccc(Cl)cc21)N1CCN(C2CCOC2)CC1', 'CC(NC(=O)Cc1ccc(-n2cnnn2)cc1)(C(=O)Nc1cncc2ccccc12)c1ccc(Cl)c(Cl)c1', 'Cc1ccncc1C(C(=O)NCCc1cccc(F)c1)N(C(=O)c1ccco1)c1ccc(C(C)(C)C)cc1', 'O=C(NCCc1cccc(F)c1)C(c1cccnc1)N(C(=O)c1ccco1)c1ccc(C2CCCC2)cc1', 'O=C(NCCc1cccc(F)c1)C(c1cccnc1)N(Cc1ccccc1C1CCC1)C(=O)c1ccco1', 'COc1ncncc1C(C(=O)NCCc1cccc(F)c1)N(C(=O)c1cocn1)c1ccc(C(C)(C)C)cc1', 'O=C(CC1(C(=O)Nc2cncc3ccccc23)CCOc2ccc(Cl)cc21)N1CCN(CC(F)(F)CO)CC1', 'CC(C)(C)c1ccc(N(C(=O)c2cocn2)C(C(=O)NCC(=O)OCc2ccccc2)c2cccnc2)cc1', 'CCOC(=O)c1ncoc1CNC(=O)C(c1cccnc1)N(C(=O)c1cocn1)c1ccc(C(C)(C)C)cc1', 'O=C(NCCc1cccc(F)c1)C(c1cccnc1)N(C(=O)c1cocn1)c1ccc(Cn2cccn2)cc1', 'O=C(NCCc1cccc(F)c1)C(c1cccnc1)N(C(=O)c1cocn1)c1ccc(OC2CCOC2)cc1'], # a list of SMILES can be provided
        "specific_parameters": None            # not required; note, this is "null" in JSON
    }]
}
configuration["parameters"]["scoring_function"] = scoring_function

In [13]:

# write the configuration file to the disc
configuration_JSON_path = os.path.join(output_dir_RL, "RL_config.json")
with open(configuration_JSON_path, 'w') as f:
    json.dump(configuration, f, indent=4, sort_keys=True)

Run REINVENT Now it is time to execute REINVENT locally.

execute REINVENT¶

python /input.py .json

In [14]:

%%capture captured_err_stream --no-stderr

# execute REINVENT from the command-line
!python {reinvent_dir}/input.py {configuration_JSON_path}

Sampling¶

In [38]:

# initialize the dictionary
configuration = {
    "version": 2,                          # we are going to use REINVENT's newest release
    "run_type": "sampling"                 # other run types: "scoring", "validation",
                                           #                  "transfer_learning",
                                           #                  "reinforcement_learning" and
                                           #                  "create_model"
}

In [39]:

# add block to specify whether to run locally or not and
# where to store the results and logging
configuration["logging"] = {
    "sender": "http://127.0.0.1",          # only relevant if "recipient" is set to "remote"
    "recipient": "local",                  # either to local logging or use a remote REST-interface
    "logging_path": os.path.join(output_dir,"sampling/progress.log"), # where the run's output is stored
    "job_name": "Sample_Mpro",       # set an arbitrary job name for identification
    "job_id": "Sample_Mpro_1"                       # only relevant if "recipient" is set to "remote"
}

In [40]:

# set up the file path, where the generated molecules are
# deposited (as SMILES)
output_SMILES_path = os.path.join(output_dir,"sampling/sampled/sampled.smi")

# add the "parameters" block
configuration["parameters"] = {
    "model_path": os.path.join(output_dir_RL,        # path to prior or trained agent
                               "results",
                               "Agent.ckpt"),
    "output_smiles_path": output_SMILES_path,       # output path where SMILES will be deposited
    "num_smiles": 100,                             # specifies, how many molecules are to be sampled
    "batch_size": 128,                              # internal batch size; should match the one used
                                                    # during training
    "with_likelihood": False                        # also provide the log-likelihood
}

In [41]:

# write the configuration file to the disc
configuration_JSON_path = os.path.join(output_dir, "sampling/sampling_config.json")
with open(configuration_JSON_path, 'w') as f:
    json.dump(configuration, f, indent=4, sort_keys=True)

Run REINVENT Now it is time to execute REINVENT locally.

execute REINVENT¶

python /input.py .json

In [42]:

%%capture captured_err_stream --no-stderr

# execute REINVENT from the command-line
!python {reinvent_dir}/input.py {configuration_JSON_path}

Generated molecules¶

In [43]:

from rdkit import Chem
import pandas as pd
from rdkit.Chem import PandasTools
#gen_smi = Chem.SmilesMolSupplier(os.path.join(output_dir_sampling,"sampled/sampled.smi"))
df = pd.read_csv (os.path.join(output_dir_sampling,"sampled/sampled.smi"))
df.columns = ['smiles']
PandasTools.AddMoleculeColumnToFrame(df, smilesCol='smiles')
PandasTools.FrameToGridImage(df, molsPerRow=4)

Out[43]:

In [ ]: