Particle Picking with Convolution Neural Network

Update 2018-1-24 EMAN2.21

In EMAN2.2 release and later version, the neural network particle picking is built in e2boxer.py

Launch e2boxer from e2projectmanager.py, choose Neural Net in Autoboxing Methods panel. Click Good Refs in Mouse Mode panel, choose a few good particles (~10 is usually enough, having more may help). Make sure to choose the good particles over micrographs of different defocus range, and the particles are centered. Then click Bad Refs, and pick some non-particles things in micrograph, like background noise and ice contamination (N>50). Click Train in the Autoboxing Methods panel. Look at the command line output. When it says done, click Autobox or Autobox All to box particles. The Auto-boxed particles are sorted by their score. Shift-click a particle to remove it, or Control-Shift-click a particle to remove this particles and all the particles after this one.

e2boxer

Old Tutorial (deprecated)

EMAN2 daily build after 2015-11-06

The program can be found in:

This program is developed by Muyuan Chen. Please contact muyuanc@bcm.edu if you have any questions.

Here we train a stack of convolutional neural nets to recognize particles in the micrograph. The basic structure of the convolutional net used in this program is described here:

This program requires Theano, in addition to other EMAN2 dependencies. Guide to install Theano can be found here:

This program runs on GPU if the GPU environment is set up in Theano. If not, it should be able to run on CPU, but the speed may be slower. Also, some functions (not very useful at this point) will be disabled if CPU is used.

Example

We use some IP3R images as a example.

input micrograph

Making Training Set

particles training set

negative training set

Train the convolutional network

convolution training output

Due to some Theano issue, if the program is run on CPU, the second image (output from the first layer) will not display.

convolution training test on image

Box particles

boxer

boxer

A zoomed in view of the particles:

result

result

That's it~

Muyuan 2015-11-05