ComplexModeler is a computational tool using CryoREAD and DiffModeler to automatically build full protein-DNA/RNA complex structure from cryo-EM maps at 0-5A resolution.
Copyright (C) 2023 Xiao Wang, Han Zhu, Genki Terashi, Daisuke Kihara, and Purdue University.
License: GPL v3. (If you are interested in a different license, for example, for commercial use, please contact us.)
Contact: Daisuke Kihara (dkihara@purdue.edu)
For technical problems or questions, please reach to Xiao Wang (wang3702@purdue.edu).
Xiao Wang, Han Zhu, Genki Terashi & Daisuke Kihara. Protein Complex Structure Modeling with Diffusion Model and AlphaFold in cryo-EM maps.bioArxiv, 2023.
@article{wang2023DiffModeler,
title={Protein Complex Structure Modeling with Diffusion Model and AlphaFold in cryo-EM maps},
author={Xiao Wang, Han Zhu, Genki Terashi, and Daisuke Kihara},
journal={bioArxiv},
year={2023}
}
Input map+single-chain structures: https://em.kiharalab.org/algorithm/ComplexModeler
For detailed introduction and protocol, please check DiffModeler and CryoREAD
CPU: >=8 cores
Memory (RAM): >=50Gb. For maps with more than 3,000 nucleotides, memory space should be higher than 200GB if the sequence is provided.
GPU: any GPU supports CUDA with at least 12GB memory.
GPU is required for DiffModeler and CryoREAD.
1. Install git
git clone --recurse-submodules https://github.com/kiharalab/ComplexModeler && cd ComplexModeler
Install anaconda from https://www.anaconda.com/download#downloads.
Then create the environment via
conda env create -f environment.yml
Each time when you want to run this software, simply activate the environment by
conda activate ComplexModeler
conda deactivate(If you want to exit)
Run the following command in the project direcotry
chmod 777 set_up.sh
./set_up.sh
If it fails, you can run set_up.sh line by line in command line.
Blast: Please follow the instructions in NCBI website to install Blast locally.
(Optional but highly recommended):
Phenix: https://phenix-online.org/documentation/install-setup-run.html
Coot: https://www2.mrc-lmb.cam.ac.uk/personal/pemsley/coot/
To verify phenix is correctly installed for final refinement step, please run
phenix.real_space_refine -h
To veryify coot is correctly installed for final refinement step, please run
coot
If it can print out the help information of this function, then the refinemnt step of our program can be supported.
If not, please always remove --refine command line in all the commands, then ComplexModeler will build structure without refinement.
Pymol (for structure visualization): https://pymol.org/2/
Chimera (for map visualization): https://www.cgl.ucsf.edu/chimera/download.html
Command Parameters
usage: main.py [-h] -F F [-P P] [--resolution RESOLUTION] [--gpu GPU] [--output OUTPUT]
[--contour CONTOUR] [--refine] [--gpu_only]
optional arguments:
-h, --help show this help message and exit
-F F input map path
-P P input fasta path
--resolution RESOLUTION
resolution for diffusion and structure refinement
--gpu GPU specify the gpu we will use
--output OUTPUT Output directory
--contour CONTOUR Contour level for input map, suggested 0.5*[author_contour]. (Float), Default
value: 0.0
--refine Optional Input. Do the last step refinement or not (Suggested to set as True).
--gpu_only only run GPU related part, server use only
Complex Structure Modeling
python3 main.py -F=[Map_Path] -P=[Fasta_Path] --contour=[contour_level] --gpu=[GPU_ID] --output=[Output_Directory] --resolution=[Map_Resolution]
[Map_Path] is the path of the experimental cryo-EM map
[Fasta_Path] is the path of the input fasta file about sequence information. The sequence information of protein is required but can be partial, DNA/RNA sequence information is optional
[contour_level] is the contour_level (suggested by author) to remove outside regions to save processing time. This is absolute density threshold, not standard deviation. If you are not sure, just set 0. Our model will automatically detect useful regions.
[GPU_ID] specifies the gpu used for inference
[Output_Directory] specifies the directory you want to save the output. If you don't specify, the default will be "Predict_Result/[map_name]". The final structure is kept as ComplexModeler.cif in this directory.
[Map_Resolution] is the resolution of the deposited maps, which is for refinement usage.
Example of fasta file
>A,B,C,D
MATPAGRRASETERLLTPNPGYGTQVGTSPAPTTPTEEEDLRR
>E,F
VVTFREENTIAFRHLFLLGYSDGSDDTFAAYTQEQLYQ
For ID line, please only include the chain id without any other information. If multiple chains include the identical sequences, please use comma "," to split different chains.
In this example, we have 6 chains in total, with A,B,C,D share the identical sequences and E,F share another identical sequences.
Here sequence information of protein is required but can be partial, DNA/RNA sequence information is optional.
If you have successfully installed phenix and coot, please also specify --refine in the command line to refine structures.
python3 main.py -F=example/20031.mrc -P=example/20031.fasta --contour=0.6 --gpu=0 --output=output_21051 --resolution=3.7
The example dir should be set up by set_up.sh. If not, please download examples here and put into example directory.
The automatically build atomic structure is saved in output_21051/Complex_Modeler.cif.
If you have successfully installed phenix and coot, please also specify --refine in the command line to refine structures.