We make all the executables, python source codes, and prebuilt databases for the Therapeutic Peptide Design Database (TP-DB, https://dyn.life.nthu.edu.tw/design) available through GitHub under the Apache License, Version 2.0.
TP-DB is generally running on a Linux system in NTHU. The local version is best suited for Linux and Mac users; For Window users, PowerShell has to be used after assuring that a 10GB memory and 20GB hard disk space are made available in your local PC.
Please note that, for a general use case, there is no need to directly download any of the files located in the links above. Those are listed for transparency. All downloads are directly handled in step 3 below, which contains a command that handles the downloading (of a 10 GB docker image the first time the command is run) as well as its execution. When step 3 is run for the first time, the docker image has to be downloaded before the code execution. From the second query onward (regardless of the query), the already downloaded docker image is loaded from users’ local computers and performs the search with results returned in a few minutes (which is slower than our webserver, see step 3 below for the reasons).
Docker makes it possible to share prebuilt computation environment for a software so that all users can have all the required dependencies and libraries running in their local computers. For this reason, we provide a docker image for TP-DB (emmanuelsalawu/tp-db). Although the TP-DB is made available through Docker Hub (https://hub.docker.com/r/emmanuelsalawu/tp-db), we strongly recommend that users follow step 3 below for the downloading and the execution of the docker image. On the other hand, if desired, a user can extract/export all the codes and the databases from the docker image. We make the relevant contents of the docker image available at http://dx.doi.org/10.5281/zenodo.5653287 on Zenodo for reference.
The easiest way for a general user to use the Therapeutic Peptide Design Database (TP-DB) is through the version we host online at https://dyn.life.nthu.edu.tw/design. Nonetheless, all the results of the version we are hosting can be reproduced and even extended through the executables, python source codes, and prebuilt-databases.
To use TP-DB on your local machine, we recommend using docker. The steps are as follows.
You would need a Mac, a Linux computer or a Windows with at least 10 GB RAM, and 20 GB free disk space in which the docker image containing the database files can be saved. This version of TP-DB has been tested on Mac OS (version 10.14.6), Windows (version 10), and Linux (Ubuntu 20.04) computers. For the best experience, Linux system is recommended. For Windows users, PowerShell has to be used.
Docker can be downloaded and installed by following the guidelines available on their website at https://docs.docker.com/get-started/. Specific download links are also provided below.
[a] Download Docker Desktop for Mac: https://desktop.docker.com/mac/stable/Docker.dmg
[b] Download Docker Desktop for Windows: https://desktop.docker.com/win/stable/Docker%20Desktop%20Installer.exe
[c] Install Docker Engine on Linux: https://docs.docker.com/engine/install/
Please note that the processing of each query is much faster when using the version hosted on our server (https://dyn.life.nthu.edu.tw/design) because the databases have been preloaded in the server’s memory and the server responds to the queries that arrive. On the other hand, the local version running through docker require to reload the databases each time, making it much slower than the webserver we provide.
Go to your computer’s terminal and execute the following. (The appropriate terminal on a Windows O/S could be the PowerShell that is running with an administrator privilege, and would not require “sudo” at the beginning of the command.)
sudo docker run -v "/tp-db:/tp-db" emmanuelsalawu/tp-db /bin/bash -c "./loadDbAndListenJuly22.exe &> /dev/null & sleep 10 && /usr/local/miniconda2/bin/python query.py --query 'A/E 0,4 D 2,3 E 2,3 H' --output /tp-db/result; echo 'Done. Now exiting'; exit "
NOTE: The first time you run TP-DB using the command, a docker image would be downloaded. This may require about 20GB (for a first-time user) data transfer through the internet. Depending on your internet connection, this may take more than 30 minutes, including the time for the download (longer) and execution (much shorter). However, subsequent running of the program with different queries makes use of the already downloaded TP-DB docker image and would run faster finishing between 2 to 5 minutes depending on your system’s configuration.
Breakdowns of what step [3] Does
The command above has the following components. You will need to change the parts in bold accordingly.
sudo docker run
-v "/tp-db:/tp-db"
emmanuelsalawu/tp-db
/bin/bash -c "./loadDbAndListenJuly22.exe &> /dev/null & sleep 10 &&
/usr/local/miniconda2/bin/python query.py
--query 'A/E 0,4 D 2,3 E 2,3 H'
--output /tp-db/result;
echo 'Done. Now exiting'; exit "
sudo docker run
This tells Docker engine that we want to run an instance.
-v "/tp-db:/tp-db"
The first /tp-db is the location (i.e. a directory) in your computer where the output will be written. It has to be a full path (and not a relative path). The second /tp-db is a location inside the docker instance. The command tells docker engine to mount a path/directory on your computer, /tp-db, to the docker instance and it should be visible as /tp-db within the docker instance. Through this, we can easily write outputs from the docker instance into a location that is accessible outside of the docker instance and can be opened on the host computer.
emmanuelsalawu/tp-db
This is the name of the TP-DB docker image that is available in docker hub https://hub.docker.com/repository/docker/emmanuelsalawu/tp-db. It is downloaded the first time we run the command and remains available locally henceforth.
/bin/bash -c "./loadDbAndListenJuly22.exe &> /dev/null & sleep 10 &&
This loads the TP-DB database and listens for incoming query. It waits for a few seconds to ensure that everything is loaded before accepting the query.
/usr/local/miniconda2/bin/python query.py
This submits the query.
--query 'A/E 0,4 D 2,3 E 2,3 H'
'A/E 0,4 D 2,3 E 2,3 H' is the example query.
--output /tp-db/result;
This is where the results would be written.
By default, the output file will be named as "result_table.html" in /tp-db folder. Users can change the name of the output table by replacing the red part. For example, changing the "—output /tp-db/mycase". The output table, "mycase_table.html", will be genrated in /tp-db folder.
echo 'Done. Now exiting';
This informs the user that the computations have finished.
exit"
This closes the docker instance and frees the resources.
Navigate to the folder you specified in step 3. Open the html file in the folder. It will contain the results identical to the one that can be obtained by using the online TP-DB. Here https://dyn.life.nthu.edu.tw/design/result?JobID=618a8525a is an example of the results page.
When a user does not need TP-DB anymore, it is possible to delete TP-DB’s docker image and free up the used disk space by executing the following command.
sudo docker image rm -f emmanuelsalawu/tp-db
NOTE: Any subsequent use of TP-DB after this step will make docker to redownload TP-DP’s docker image as explained in step 3 above.
Python source codes for TP-DB are available through GitHub: https://github.com/emmanuelsalawu/tp-db where users can go through the codes and, if desired, copy and modify it to suit their needs. The data comprising the database can be found at http://dx.doi.org/10.5281/zenodo.4628579 (e.g. see “sequences_helix_corrected_dec292017.fasta.zip”).