Skip to content
forked from thuem/THUNDER

A particle-filter framework for robust cryoEM 3D reconstruction

License

Notifications You must be signed in to change notification settings

maobaoqi/THUNDER

 
 

Repository files navigation

THUNDER

A particle-filter framework for robust cryoEM 3D reconstruction

Literature

bioRvix, A particle-filter framework for robust cryoEM 3D reconstruction

Citation

Bai R, Wan R, Yan C, et al. Structures of the fully assembled Saccharomyces cerevisiae spliceosome before activation[J]. Science, 2018:eaau0325.

Release Note

  • Version 1.4.11: This release fixes a bug which occurs during the initialisation step of a few datasets.
  • Version 1.4.10: This release fixes a compilation bug which occurs when CUDA version is below 9.0.
  • Version 1.4.9: Graphic User Inference (GUI) thunder_stackview is released. It is used for viewing the result of 2D classification and selecting desired particles.
  • Version 1.4.8: This release elevates the support of NVIDIA GeForce GTX 1080/1080Ti.
  • Version 1.4.7: This release fixes minor bugs.
  • Version 1.4.6: This release fixes a bug occurring during applying mask on the reference.
  • Version 1.4.5: This release fixes a bug which appears when 8 or more GPUs are used in a single workstation. This release also fixes some minor bugs.
  • Version 1.4.4: This release elevates the support of NVIDIA V100.
  • Version 1.4.3: This release increases the stability of 2D classification.
  • Version 1.4.2: This release enables THUNDER to run on single workstation with one or multiple GPUs.
  • Version 1.4.1: This release fixes minor bugs.
  • Version 1.4.0: This release enables THUNDER to run on GPU clusters.

Quick Start

Installation from Source

Installation from source requires CMake and a C/C++ compiler with MPI wrapper.

git clone [email protected]:thuem/THUNDER.git THUNDER
cd THUNDER
mkdir build
cd build
cmake ..
make
make install

Installation of specified version of THUNDER, such as double-precision, designed SIMD version or GPU version, is described in the manual in manual folder.

Dependency of GPU Version

CUDA 8.0 or above and NCCL2 are required. You may download CUDA from https://developer.nvidia.com/cuda-toolkit and NCCL2 from https://developer.nvidia.com/nccl.

Please make sure that the proper version of NCCL2 is installed, as it depends on the version of CUDA, operating system and computer architecture.

Running THUNDER

Please view the manual in manual folder.

GUI Installation

Installation of GUI requires Qt5. You may download an open source version Qt5 from https://www1.qt.io/download-open-source-access/. You may also install Qt5 with yum or apt-get.

cd THUNDER/gui/thunder_stackview
mkdir build
cd build
qmake ..
make
make install

Viewing Result of 2D Classification and Selecting Desired Particles

You may type thunder_stackview to get help.

If you want simply view result of 2D classification, you may type thunder_stackview Reference_Round_XXX.mrcs or thunder_stackview Reference_Final.mrcs. Moreover, if you want to select desired particles, please type thunder_stackview Reference_Round_XXX.mrcs -thu Meta_Round_XXX.thu or thunder_stackview Reference_Final.mrcs -thu Meta_Final.thu. After selecting the classes you desired, please use the Save thu button to save the selection.

Authors

See AUTHORS.txt file.

License

See LICENSE.txt file for details.

Acknowledgements

This work was supported by funds from The National Key Research and Development Program, National Natural Science Foundation of China, Advanced Innovation Center for Structural Biology, Tsinghua-Peking Joint Center for Life Sciences and One-Thousand Talent Program by the State Council of China. We acknowledge the National Supercomputing Center in Wuxi and Tsinghua University Branch of China National Center for Protein Sciences Beijing for providing facility supports in computation.

About

A particle-filter framework for robust cryoEM 3D reconstruction

Resources

License

Stars

Watchers

Forks

Packages

No packages published

Languages

  • C++ 57.4%
  • Cuda 35.8%
  • C 4.4%
  • TeX 1.3%
  • CMake 0.9%
  • Python 0.2%