Currently, ST-Analyzer should be installed through Conda/Anaconda. Conda's installation instructions begin here.
ST-Analyzer has been tested on macOS (x64 and arm64), Ubuntu (x64), and WSL (x64, running Ubuntu).
See the section below for regular installation.
For development installation, skip to the section Installation via Conda (ST-Analyzer developers).
ST-Analyzer works best when it is installed in a fresh conda environment using the conda-forge channel. The example below installs into a new environment named sta:
conda create -n sta -c conda-forge stanalyzerTo use ST-Analyzer, first activate its environment:
conda activate sta # required once per shell sessionThen, you should have two new commands available: stanalyzer and sta-server. Their usage is explained in the Usage section.
Some analysis programs require packages that do not work on all platforms:
| Analysis | Requires |
|---|---|
| Secondary Structure | dssp |
| Pore Radius Histogram | hole2 |
| SASA | freesasa |
If you don't need these programs, then you don't need to do anything.
Otherwise, you can try to install them:
conda install -c conda-forge dssp hole2 freesasaIf installation fails, you may want to contact those packages' developers.
If you get a message like "environment already exists", you should either use a different name or remove the old one. This command removes the sta environment:
# WARNING: Removes ST-Analyzer's environment (or other w/ name sta)
conda env remove -n staDevelopment installation entails obtaining the ST-Analyzer source code and linking it to your python environment. The helper script dev_install.sh can run most of the steps for you:
# install ST-Analyzer's dependencies in a new environment
conda create -n sta-dev --only-deps -c conda-forge nk53::stanalyzer
# enter environment
conda activate sta-dev
# download source code if you haven't already; otherwise skip this step
git clone "git@github:nk53/stanalyzer.git"
# go to project directory
cd stanalyzer
# install package in editable mode
pip install -e .Editable installations let you update the source code without needing to repackage/reinstall anything; your changes will be reflected immediately. The only reason to mess with conda is if something in the upstream environment changes, in which case conda update --only-deps stanalyzer should be enough.
You'll only need these if you want static type checking or to compile the online documentation. Check dev_requirements.txt if either of those interest you. You can pick and choose, or if you want all of the listed packages, do this:
conda install -c conda-forge --file dev_requirements.txtIf you are testing recent development changes to static files such as forms.js or style.css, you browser may be serving you the old version of a file. Some ways you can get around this:
- Clear your browser's cache of this site
- Disable caching for this site
- Open the site in a new private browsing window
- If using Firefox with developer tools enabled, open the network tab and toggle the "Disable caching" option.
ST-Analyzer is in alpha development stage; it is intended to be run and accessed on your personal computer.
If ST-Analyzer was installed successfully, you should be able to start the Web GUI with 1-2 commands, run from anywhere:
conda activate sta # if not already active
sta-serverThis runs the server process in the foreground of your current shell.
If you did not receive any error messages, then the server should be listening for localhost connections, and will show the corresponding URL. At the time of this writing, it should be http://127.0.0.1:8000. You can copy and paste that into your browser's address bar to reach the ST-Analyzer front page.
Advanced users often want to automate analysis themselves without dealing with the web UI. ST-Analyzer was designed with this goal in mind. All analysis processes can be run from the command line directly. In fact, this is what the web GUI is doing behind the scenes.
The easiest way to get started with the command-line interface (CLI) is to first use the web interface to create a project and run a small sample analysis, e.g., the "System size" analysis with default settings. When an analysis job is successfully submitted through the web GUI, it writes the exact command needed to run the analysis from the command-line. E.g.:
args: stanalyzer system_size --time_step 1 --out system_size.outIn all cases, the web GUI uses the verbose version of a command's options. To see all available options for a command and their aliases, use stanalyzer [analysis_program] -h. E.g.:
$ stanalyzer system_size -h
usage: stanalyzer system_size [-h] [-p FILE] [-t FILE [FILE ...]] [-o OUT]
[-ts TIME_STEP] [-i INTERVAL] [-a]
options:
-h, --help show this help message and exit
-p FILE, --psf FILE File containing system topology (default: use project
settings)
-t FILE [FILE ...], --traj FILE [FILE ...]
One or more coordinate containing files (default: use
project settings)
-o OUT, --out OUT File to write results (default: stdout)
-ts TIME_STEP, --time_step TIME_STEP
Amount of time between frames in trajectory files
-i INTERVAL, --interval INTERVAL
step size when reading frames (default: 1 = read every
frame)
-a, --include-anglesconda remove stanalyzerNote that pip's editable installation produces stanalyzer.egg-info/. This directory can interfere with complete package removal. To ensure that __editable__.stanalyzer-*.pth and stanalyzer-*.dist-info/ can be removed from your distribution's site-packages, you should either ensure you are not in the same directory as stanalyzer.egg-info/ or remove it.
Then simply use pip uninstall.
# either this
cd && pip uninstall stanalyzer
# or this
rm -r stanalyzer.egg-info && pip uninstall stanalyzer