DMcloudComplex

DMcloudComplex is a macromolecular structure modeling tool using point cloud matching with diffusion model and 3D U-net (CryoREAD) for cryo-EM maps at 3-10A resolution.

Installation

Clone the Repository

First, clone the DMCloud repository and the submodules:

git clone https://github.com/kiharalab/DMcloudComplex.git
cd DMcloudComplex
git submodule update --init --recursive

Download Model Weights for Diffusion Model

Download the pre-trained diffusion model weights:

cd DMCloud
mkdir -p best_model
wget -O best_model/diffusion_best.pth.tar https://github.com/kiharalab/DMCloud/releases/download/weights/diffusion_best.pth.tar

Prerequisites

First, install a conda package manager if you don't have one:

• Miniforge (recommended): Download from https://github.com/conda-forge/miniforge
• Miniconda: Download from https://docs.conda.io/en/latest/miniconda.html

Environment Setup

Using conda:

conda env create -f environment.yml
conda activate dmcloud
pip install ortools pulp numba_progress progress

Or using mamba (faster alternative):

mamba env create -f environment.yml
mamba activate dmcloud
pip install ortools pulp numba_progress progress

Note: If you installed Miniforge, mamba is included by default. For Miniconda users, install mamba with: conda install mamba -n base -c conda-forge

Usage

python DMcloudComplex.py --map [MAP file] --gpu [GPU-ID] --output [OutPut Dir] --contour_nuc [Contour Level for DNA/RNA] --contour_prot [COntour Level for Protein] --protein [Protein Models] --nucleic [DNA/RNA models] [options]

Example of Protein-RNA complex modeling

Command

python ./DMcloudComplex.py --map ./examples/21051.mrc --gpu 0 --output ./OutPut --contour_nuc 0.3 --contour_prot 0.3 --protein examples/A.pdb,examples/B-C.pdb --nucleic examples/TEMP_6v5c.cif

Expected Output files

DNA/RNA models in CIF/PDB format

best_rna_models.cif
best_rna_models.pdb

Protein models in CIF/PDB format

best_protein_models.cif
best_protein_models.pdb

Enhanced Map

DMcloudProtein_run0/Diffusion/diffusion.mrc

CryoREAD Outputs (Point cloud)

CryoREAD_run/graph_atomic_modeling/pho_LDP/pho_LDPdens.cif
CryoREAD_run/graph_atomic_modeling/sugar_LDP/sugar_LDPdens.cif

CryoREAD Outputs (Probability Map)

CryoREAD_run/graph_atomic_modeling/pho_LDP/pho_LDP.mrc
CryoREAD_run/graph_atomic_modeling/sugar_LDP/sugar_LDP.mrc

Parameters

Parameter	Type	Default	Required	Description
--map	str	-	Yes	Input MAP file in MRC/MAP format
--protein	str	-	Yes	Protein Model CfoDMcloudComplex. Multiple file names sould be separated by ",", like "AAA.pdb,BBB.pdb"
--nucleic	str	-	Yes	DNA/RNA Model in CIF/PDB format. Multiple file names sould be separated by ",", like "RNA2.cif,RNA1.cif"
--output	str	./Predictions/	No	Final output directory
--contour_nuc	float	0.0	No	Contour Level for DNA/RNA regions
--contour_prot	float	0.0	No	Contour Level for Protein regions
--gpu	int	0	No	GPU-ID

Important Notes

• Config File Path Resolution: When using MRC/MAP files as targets, DMcloud automatically uses config/diffusion.json as the default configuration. The config path is resolved relative to the DMcloud.py script location, so the tool works correctly when called from any directory.
• Cross-Directory Usage: You can run DMcloud from any directory, and it will automatically find the default config file relative to its installation location.

License

License: GPL v3. (If you are interested in a different license, for example, for commercial use, please contact us.)

Contact

Contact: Genki Terashi (gterashi@purdue.edu) Daisuke Kihara (dkihara@purdue.edu)

---