Combs2024
Combs repo associated with Lu et al 2024 Science paper
Install / Use
/learn @npolizzi/Combs2024README
Combs2: Convergent motifs for binding sites 2.0
SYNOPSIS +++++++++
Combs2 is an open-source Python package for interfacial protein design.
Combs generates and uses van der Mers (vdMs) to analyze and design
interaction motifs in proteins, in small-molecule-protein interfaces as
well as larger protein-protein interfaces.
INSTALLATION ++++++++++++ Install Combs2 by cloning the github repository, e.g. ::
> git clone https://github.com/npolizzi/Combs2024.git
Note that you can keep Combs2 up-to-date (useful for bug-fixes and updates to the code) by: ::
> git pull https://github.com/npolizzi/Combs2024.git
Within Combs2/docs/ use the env_combs_(platform).yml file to create a conda virtual environment for Combs2 by, e.g. ::
> conda env create -f env_combs_mac.yml
You will need to update the prefix: line at the bottom of the .yml file to point to your miniconda installation directory.
Change the path up to .../miniconda3/ the environment directories will be created for you ::
This will install all the necessary packages on which Combs2 depends but requires miniconda to be pre-installed.
To install miniconda for Mac, go to the Anaconda website <https://docs.conda.io/projects/conda/en/latest/user-guide/install/macos.html>_ and follow the instructions.
If you are experiencing problems installing the environment, a minimal conda environment that should get you going can be created by ::
> conda create --name env_combs python=3.8
> conda activate env_combs
> pip install sklearn scipy pandas numba numpy prody pyarrow jupyter matplotlib
If you are installing on a Mac M1, numba is not yet pip-installable. Try this workaround: ::
> pip install -i https://pypi.anaconda.org/numba/label/wheels_experimental_m1/simple llvmlite
> pip install "numpy==1.21.*" importlib-metadata
> pip install -i https://pypi.anaconda.org/numba/label/wheels_experimental_m1/simple numba
Note that, for some older Mac operating systems, you may need to specify scipy==1.1.0 (If you see a Segmentation Fault when trying to use Combs2, try this.). After creating the env_combs environment, add the path to Combs2 to your .bash_profile or .bashrc (which is usually in your home directory): ::
> vi .bash_profile
export PYTHONPATH="/Users/npolizzi/Projects/design/Combs2024/:$PYTHONPATH"
Exchange the python path above to your own path to Combs2. For Linux, you may need to drop the quotes or change them to single quotes. Now you can activate the Combs2 virtual environment and use Combs2, interactively or in scripts, e.g. ::
> conda activate env_combs
> ipython
Then, in an IPython session or Juptyer notebook:
.. code-block:: python
import combs2
Note that you should activate the conda environment before using Combs. Otherwise you will likely encounter missing-package errors.
DATABASE ++++++++
The van der Mer databases are currently hosted on DropBox (subject to change) at this link <https://www.dropbox.com/sh/a5wakk7nonc03bv/AACbar6bDBua-HH7L_-2iO-0a?dl=0>_. The zipped file is about 7.5 GB currently. For a list of available chemical groups and the amino acids from which they derive, see CG_dicts.txt in Combs2024/combs2/files/. Download and unzip the database file by: ::
> tar -xvf database.tar.gz
You do not need to unzip the gzipped files within the vdMs folder.
SUPPORTING SOFTWARE
+++++++++++++++++++
Some of the Combs functionality (e.g. classifying rotamers or secondary structure) currently depends on third-party software such as DSSP and MolProbity. You are encouraged to download and install DSSP <https://swift.cmbi.umcn.nl/gv/dssp/>_ (and also point ProDy to the DSSP installation path if necessary). You are highly encouraged to download and install MolProbity from the Richardson Lab's GitHub page <https://github.com/rlabduke/MolProbity>_. Some functions within Combs require a path to Probe, Reduce, or Rotalyze within MolProbity.
TUTORIAL
++++++++
Update: See the script Combs2024/tutorials/HPC_scripts/biotin/run_design.py for an example of the most up-to-date usage of COMBS.
A link to a couple Zoom recordings for combs tutorials can be found here <https://www.dropbox.com/scl/fo/u10ohdc7bjwbjv8xkniwg/AOigN8huLcb_tktWmjJlnoQ?rlkey=jaslipjhh6qfddmxsk9q484tn&dl=0>_. This covers basic file creation and some background re: vdMs.
DOCUMENTATION +++++++++++++ Very little unfortunately, see tutorial and example scripts.
LICENSE +++++++
Combs2 is available under MIT License. See LICENSE.txt for more details.
Related Skills
node-connect
347.2kDiagnose OpenClaw node connection and pairing failures for Android, iOS, and macOS companion apps
frontend-design
108.0kCreate distinctive, production-grade frontend interfaces with high design quality. Use this skill when the user asks to build web components, pages, or applications. Generates creative, polished code that avoids generic AI aesthetics.
openai-whisper-api
347.2kTranscribe audio via OpenAI Audio Transcriptions API (Whisper).
qqbot-media
347.2kQQBot 富媒体收发能力。使用 <qqmedia> 标签,系统根据文件扩展名自动识别类型(图片/语音/视频/文件)。
