Home | Resources | ROME System


DeepEM: a deep-learning based particle recognition program

Photo of ML SystemA deep learning-based algorithmic framework, DeepEM, is developed for single-particle recognition from noisy cryo-EM micrographs, enabling automated particle picking, selection and verification in an integrated fashion. The kernel of DeepEM is built upon a convolutional neural network of eight layers, which can be recursively trained to be highly “knowledgeable”. Our approach exhibits improved performance and high precision when tested on the standard KLH dataset. DeepEM is currently experimented with Matlab 2014b. DeepEM is expected to be introduced into ROME 2.0 as a Matlab-independent, parallel computing module in the near future.

Further Reading

Y. Zhu, Q. Ouyang, Y. Mao. A deep convolutional neural network approach to single-particle recognition in cryo-electron microscopy. BMC Bioinformatics 18, 348 (2017). Learn More. arXiv: 1605.05543.

Software Download

Click here: Matlab source code. Please note that this code was only tested on Matlab 2014b, and it is known that the later version of Matlab has some compatibility issues.