ContainersNVIDIADeepSeek-R1NVIDIALlama-3.1-Nemotron-70B-InstructNVIDIAPyTorch
CollectionsNVIDIAOmniverse Kit (FB)NVIDIADeepStream SDKNVIDIAOmniverse Kit App Streaming
ModelsNVIDIAStyleGAN3 pretrained modelsNVIDIAPeopleNetNVIDIATrafficCamNet
ResourcesNVIDIARiva Skills Quick Start
Helm ChartsNVIDIAGPU OperatorNVIDIANVIDIA NIM Operator
Scripps
AutoDock-GPU
Container
Scripps
AutoDock-GPU

The AutoDock Suite is a growing collection of methods for computational docking and virtual screening, for use in structure-based drug discovery and exploration of the basic mechanisms of biomolecular structure and function.

AutoDock-GPU

The AutoDock-GPU Suite is a growing collection of methods for computational docking and virtual screening, for use in structure-based drug discovery and exploration of the basic mechanisms of biomolecular structure and function. More info on AutoDock-GPU be located at https://ccsb.scripps.edu/autodock/ and official github page

See here for a document describing prerequisites and setup steps for all HPC containers and instructions for pulling NGC containers.

System requirements

Before running the NGC AutoDock-GPU container please ensure your system meets the following requirements.

x86_64

  • Pascal(sm60), Volta(sm70), or Ampere (sm80) NVIDIA GPU(s)
  • CUDA driver version >= r450, -or- r418, -or- r440

arm64

  • Pascal(sm60), Volta(sm70), or Ampere (sm80) NVIDIA GPU(s)
  • CUDA driver version >= r450

Running AutoDock-GPU

Executables

autodock_gpu_128wi: primary autodock executable

Command invocation

An example command is:

autodock_gpu_128wi -ffile protein.maps.fld -lfile rand-example.pdbqt -nrun 100 -lsmet ad -resnam CUDAout

Examples

The following examples demonstrate how to run the NGC AutoDock-GPU container under the supported runtimes.

Running with docker

Once authenticated the following modes of running are supported:

Command line execution with docker

In order to run the water benchmark: download and permission, the dataset from the AD-GPU_set_of_42. Copy and paste the following script, called check.py into your working directory.

#!/usr/bin/env python3
import csv
import glob
import os
import re
import subprocess
import sys


def run_all(executable):
  dirlist = glob.glob("data/*")
  for d in dirlist:
      os.chdir(d)
      # There's potentially 3 runs to do here, doing 1 to save time.
      for i in range(1):
          subprocess.run("{} -ffile protein.maps.fld -lfile rand-{}.pdbqt -nrun 100 -lsmet ad -heuristics 1 -autostop 1 -maxnev 8000000 -resnam CUDAout-{}".format(executable, i, i), shell=True).check_returncode()
      os.chdir("../..")


def process_file(filename):
    re_ligandname = re.compile(r'^Ligand name: (.+)')
    re_intermolecular = re.compile(r'^(?=.*\bIntermolecular\b).*=(.*)')
    re_internal = re.compile(r'^(?=.*\bInternal\b).*=(.*)')
    ligand = ''
    curr_energy = sys.float_info.max
    best_energy = sys.float_info.max
    with open(filename) as f:
        for line in f:
            match = re_ligandname.search(line)
            if match:
                ligand = match.group(0).split()[2]
                continue
            match = re_intermolecular.search(line)
            if match:
                curr_energy = float(match.group(1).split()[0])
                continue
            match = re_internal.search(line)
            if match:
                curr_energy = curr_energy + float(match.group(1).split()[0])
                if (curr_energy < best_energy):
                    best_energy = curr_energy
                continue
    return best_energy


def check_output(reference_energies):
    results = {}
    err = 0.0
    cnt = 0
    for key, val in reference_energies.items():
        for dlgfile in glob.glob('data/*/CUDAout-*.dlg'):
           best_energy = process_file(dlgfile)
           diff = val - best_energy
           err = err + diff
           cnt = cnt + 1
           try:
               results[key].append(diff)
           except KeyError:
               results[key] = [diff]
    return (err/cnt)


reference_energies = {}
with open('ligand_properties.csv') as f:
    reader = csv.reader(f)
    cnt = 0
    for row in reader:
        cnt = cnt + 1
        if cnt == 1:
            continue
        # There are some bad ligands for which we have no reference energy
        if row[4] == '':
            continue
        reference_energies[row[1]] = float(row[4])

for numwi in [128]:
    run_all("autodock_gpu_{}wi".format(numwi))
    meandiff = check_output(reference_energies)
    print('{} : Mean Error: {:.3f}'.format(numwi, meandiff))
    if meandiff > 0.5:
        exit(1)

print("Checks passed.")


$ git clone https://github.com/diogomart/AD-GPU_set_of_42.git
$ gunzip AD-GPU_set_of_42/data/*/*map.gz
$ docker run -ti --gpus all -v :workdir --workdir /workdir nvcr.io/hpc/autodock:2020.06 sh -c "./check.py"

Example of successful AutoDock-GPU output:

AutoDock-GPU version: 09773678fc7e39677061d765b767f4bae8930fb7-dirty
CUDA Setup time 2.031275s
(Thread 0 is setting up Job 0)
Running Job #0:
Using heuristics: -lsmet sw -nev 7468504
Local-search chosen method is: swSolis-Wets (sw)
Rest of Setup time 0.037800s
Executing docking runs, stopping automatically after either reaching 0.15 kcal/mol standard deviation of
the best molecules of the last 4 * 5 generations, 27000 generations, or 7468504 evaluations:
Generations |  Evaluations |     Threshold    |  Average energy of best 10%  | Samples |    Best energy
------------+--------------+------------------+------------------------------+---------+-------------------
          0 |          150 |   -2.91 kcal/mol |   -4.10 +/-    0.40 kcal/mol |       5 |   -4.82 kcal/mol
          5 |       125155 |   -2.91 kcal/mol |   -5.70 +/-    1.40 kcal/mol |    8066 |  -11.91 kcal/mol
         10 |       250790 |   -5.69 kcal/mol |   -9.55 +/-    0.83 kcal/mol |     490 |  -12.35 kcal/mol
         15 |       377698 |   -9.53 kcal/mol |  -11.80 +/-    0.42 kcal/mol |      52 |  -12.62 kcal/mol
         20 |       505556 |  -11.68 kcal/mol |  -12.42 +/-    0.18 kcal/mol |      14 |  -12.67 kcal/mol
         25 |       633659 |  -12.23 kcal/mol |  -12.48 +/-    0.19 kcal/mol |      16 |  -12.74 kcal/mol
         30 |       762773 |  -12.30 kcal/mol |  -12.59 +/-    0.14 kcal/mol |      16 |  -12.83 kcal/mol
         35 |       892173 |  -12.46 kcal/mol |  -12.73 +/-    0.15 kcal/mol |      14 |  -13.02 kcal/mol
         40 |      1021489 |  -12.57 kcal/mol |  -12.81 +/-    0.13 kcal/mol |      14 |  -13.05 kcal/mol
         45 |      1151059 |  -12.68 kcal/mol |  -12.83 +/-    0.11 kcal/mol |      19 |  -13.06 kcal/mol
         50 |      1280984 |  -12.74 kcal/mol |  -12.84 +/-    0.10 kcal/mol |      13 |  -13.07 kcal/mol
Docking time 3.916151s
------------+--------------+------------------+------------------------------+---------+-------------------
                                   Finished evaluation after reaching
                                  -12.80 +/-    0.13 kcal/mol combined.
                                60 samples, best energy   -13.07 kcal/mol.
Shutdown time 0.000570s
Job #0 took 3.955 sec after waiting 3.373 sec for setup
(Thread 0 is processing Job 0)
Run time of entire job set (1 files): 8.413 sec
Savings from multithreading: -2.039 sec
Idle time of execution thread: 4.458 sec
All jobs ran without errors.

Interactive shell with docker

The following command will launch an interactive shell in the AutoDock-GPU container using docker mounting $HOME/data from the underlying system as /data in the container:

$ docker run -it --rm --gpus all -v $HOME/data:/data --workdir /data nvcr.io/hpc/autodock:2020.06

Where:

  • -it: start the container with an interactive terminal (short for --interactive --tty)
  • --rm: make container ephemeral (removes container on exit)
  • -v $(pwd):/host_pwd: bind mount the current working directory into the container as /host_pwd
  • --gpus all: vanilla docker GPU provisioning
  • --workdir /data: sets working directory inside the container

This should produce a root prompt within the container:

root@3a8c8b7c3a88:/data#

Running with Singularity

Pull the image

Save the NGC AutoDock-GPU container as a local Singularity image file:

$ singularity build autodock.simg docker://nvcr.io/hpc/autodock:2020.06

The AutoDock-GPU Singularity image is now saved in the current directory as autodock.simg

Note: Singularity/2.x

In order to pull NGC images with singularity version 2.x and earlier, NGC container registry authentication credentials are required.

To set your NGC container registry authentication credentials:

$ export SINGULARITY_DOCKER_USERNAME='$oauthtoken'
$ export SINGULARITY_DOCKER_PASSWORD=

More information describing how to obtain and use your NVIDIA NGC Cloud Services API key can be found here.

Note: Singularity 3.1.x - 3.2.x

There is currently a bug in Singularity 3.1.x and 3.2.x causing the LD_LIBRARY_PATH to be incorrectly set within the container environment. As a workaround The LD_LIBRARY_PATH must be unset before invoking Singularity:

$ LD_LIBRARY_PATH="" singularity exec ...

Command line execution

In order to run a test benchmark: download, permission, and run the example script from the NGC Examples Repository.

wget ...
chmod +x check.py
singularity run --nv -B ${PWD}:/host_pwd --pwd /host_pwd ./check.py

Example of successful AutoDock-GPU output:

AutoDock-GPU version: 09773678fc7e39677061d765b767f4bae8930fb7-dirty
CUDA Setup time 2.031275s
(Thread 0 is setting up Job 0)
Warning: unknown argument '-maxnev'.
Running Job #0:
Using heuristics: -lsmet sw -nev 7468504
Local-search chosen method is: swSolis-Wets (sw)
Rest of Setup time 0.037800s
Executing docking runs, stopping automatically after either reaching 0.15 kcal/mol standard deviation of
the best molecules of the last 4 * 5 generations, 27000 generations, or 7468504 evaluations:
Generations |  Evaluations |     Threshold    |  Average energy of best 10%  | Samples |    Best energy
------------+--------------+------------------+------------------------------+---------+-------------------
          0 |          150 |   -2.91 kcal/mol |   -4.10 +/-    0.40 kcal/mol |       5 |   -4.82 kcal/mol
          5 |       125155 |   -2.91 kcal/mol |   -5.70 +/-    1.40 kcal/mol |    8066 |  -11.91 kcal/mol
         10 |       250790 |   -5.69 kcal/mol |   -9.55 +/-    0.83 kcal/mol |     490 |  -12.35 kcal/mol
         15 |       377698 |   -9.53 kcal/mol |  -11.80 +/-    0.42 kcal/mol |      52 |  -12.62 kcal/mol
         20 |       505556 |  -11.68 kcal/mol |  -12.42 +/-    0.18 kcal/mol |      14 |  -12.67 kcal/mol
         25 |       633659 |  -12.23 kcal/mol |  -12.48 +/-    0.19 kcal/mol |      16 |  -12.74 kcal/mol
         30 |       762773 |  -12.30 kcal/mol |  -12.59 +/-    0.14 kcal/mol |      16 |  -12.83 kcal/mol
         35 |       892173 |  -12.46 kcal/mol |  -12.73 +/-    0.15 kcal/mol |      14 |  -13.02 kcal/mol
         40 |      1021489 |  -12.57 kcal/mol |  -12.81 +/-    0.13 kcal/mol |      14 |  -13.05 kcal/mol
         45 |      1151059 |  -12.68 kcal/mol |  -12.83 +/-    0.11 kcal/mol |      19 |  -13.06 kcal/mol
         50 |      1280984 |  -12.74 kcal/mol |  -12.84 +/-    0.10 kcal/mol |      13 |  -13.07 kcal/mol
Docking time 3.916151s
------------+--------------+------------------+------------------------------+---------+-------------------
                                   Finished evaluation after reaching
                                  -12.80 +/-    0.13 kcal/mol combined.
                                60 samples, best energy   -13.07 kcal/mol.
Shutdown time 0.000570s
Job #0 took 3.955 sec after waiting 3.373 sec for setup
(Thread 0 is processing Job 0)
Run time of entire job set (1 files): 8.413 sec
Savings from multithreading: -2.039 sec
Idle time of execution thread: 4.458 sec
All jobs ran without errors.

Interactive shell

The following command will launch an interactive shell in the AutoDock-GPU container using singularity shell:

$ singularity shell --nv -B

Where:

  • --nv: expose the host GPU(s) to the container
  • -B: bind user-defined directory into the container

This should produce a Singularity shell prompt within the container:

Singularity: Invoking an interactive shell within container...

Singularity Autodock.simg:~/example>

Suggested Reading

Autodock main page Autodock wiki Autodock officional github page

Publisher
Scripps
Latest Tag2020.06
UpdatedJune 22, 2020 UTC
Compressed Size79.01 MB
Multinode SupportNo
Multi-Arch SupportYes
Labels
arm64HPCx86_64

NVIDIA uses cookies to improve your experience on our web site. We and our third-party partners also use cookies and other tools to collect and record information you provide as well as information about your interactions with our websites for performance improvement, analytics, and to assist in marketing efforts. By clicking "Accept All", you consent to our use of cookies and other tools as described in our Cookie Policy. You can manage your cookie settings by clicking on "Manage Settings." By continuing to use this site or by clicking one of the buttons below, you agree to our Terms of Service (which contains important waivers). Please see our Privacy Policy for more information on our privacy practices.