Point region shape n x

Topaz Image Pre-processing Command Generator

Create commands for pre-processing images (downsampling and normalization) for use in this Topaz picking GUI and for use in Topaz training.

Fill in parameters and run the resulting command          Must change Sometimes changed Rarely changed


Input micrographs
Output folder
Number of CPUs
Scale factor

Advanced options
Pixel sampling
Number of iterations
Random seed


topaz preprocess --scale 8 --num-workers ## --pixel-sampling 200 --niters 100 --seed 1 --verbose -o /path/to/output/preprocessed/images/ /path/to/input/images/*.mrc

Pick Annotations Image Annotations × ↑ ↓ + −
  • Home
  • Project
    • Load
    • Save
    • Settings
    • Add local images
    • Add images from absolute paths
    • Add images from URL
    • Add urls or paths from text file
  • Particle Picks
    • Export Picks (as csv)
    • Import Picks (from csv)
    • Preview Picks
    • Download as Image
  • View
    • Toggle image grid view
    • Toggle left sidebar
    • Toggle image filename list
    • Toggle attributes editor
    • Toggle annotation editor (Space)
    • Show/hide pick boundaries (b)
    • Show/hide pick labels (l)
  • Help
Open Project Save Project Update Project Settings Import Annotations from CSV Download Annotations as CSV Switch to Image Grid View Toggle Annotation Editor Previous Image Toggle Image List Next Image Zoom In Zoom Out Select All Picks Copy Picks Paste Picks Paste Pick in Multiple Images Undo Picks Pasted in Multiple Images Delete Pick
Sorry, your browser does not support HTML5 Canvas functionality which is required for this application.
Group by 
Selected 0 of 0 images in current group, show

Settings

Project Name
Default Path
If all images in your project are saved in a single folder, set the default path to the location of this folder. The VIA application will load images from this folder by default. Note: a default path of "./" indicates that the folder containing the topaz.html application file also contains the images in this project. For example: /datasets/VOC2012/JPEGImages/ or C:\Documents\data\ (note the trailing / and \)
Search Path List
If you define multiple paths, all these folders will be searched to find images in this project. We do not recommend this approach as it is computationally expensive to search for images in multiple folders.
    Particle Pick Label
    By default, each region in an image is labelled using the region-id. Here, you can select a more descriptive labelling of regions.
    Particle Pick Label Font
    Font size and font family for showing region labels.
    Preload Buffer Size
    Images are preloaded in buffer to allow smoother navigation of next/prev images. A large buffer size may slow down the overall browser performance. To disable preloading, set buffer size to 0.
    On-image Annotation Editor
    When a single particle pick is selected, the on-image annotation editor is activated which the user may use to update annotations of this pick. By default, this on-image annotation editor is placed near the selected region.

    File Not Found

    Filename:

    We recommend that you update the default path in project settings to the folder which contains this image.

    A temporary fix is to use browser's file selector to manually locate and add this file. We do not recommend this approach because it requires you to repeat this process every time your load this project in the VIA application.

    1. Convert MRCs to PNGs: Use Topaz to convert your mrc files to PNGs so you can use this GUI to make training picks and analyze Topaz results:

        topaz preprocess /full/path/to/mrcs/*.mrc -s [binning] -o /full/path/to/input/images/ -t [number_of_CPU_threads]

    2. Pick Particles: Add images then pick particles:
      • To add images:
        • Click Add Images under 'Project' to the left, or
        • In the 'Project' menu, select Add images using absolute paths.
        • Note: This Topaz GUI accepts JPEG, PNG, and BMP files.
      • To pick particles:
        • Left-click on the center of particles for several micrographs.
        • To zoom in/out, either press Ctrl and scroll your mouse wheel on an image or press + or - (equal '=' will reset the zoom).
        • To delete a misplaced point, click the point again and press 'd'.
        • To adjust a point's location, click the point again and either use the arrow keys on your keyboard or click-and-drag to move it.
        • Note: Topaz works best if you identify 500+ particles representing all orientations of the particle and if you pick across several images. More training particles and views means more accurate Topaz picks.

    3. Export Particles: To export particles, click Particle Picks → Export Picks in the top menu. These picks may then be used in Topaz training.

    4. Analyze Topaz Picks: Load Topaz picks by clicking Particle Picks → Import Picks.

      5. Note: It is good practice to save your particles/project regularly. Do this by clicking Particle Picks → Export Picks or Project → Save in the top menu.
    Loading ...
        

    Topaz Version 0.0.0                                                             VIA Version 2.0.5

    Topaz is a pipeline for particle detection in cryo-electron microscopy images using convolutional neural networks trained from positive and unlabeled examples. The original research can be found here.

    Topaz GUI uses VGG Image Annotator (VIA), which is an image annotation tool that can be used to define regions in an image and create textual descriptions of those regions. VIA is an open source project developed at the Visual Geometry Group and released under the BSD-2 clause license.

    Here is a list of some salient features of the Topaz GUI version of VIA:

    • based solely on HTML, CSS and Javascript
    • can be used off-line (full application in a single html file of size < 450KB)
    • only requires a modern web browser (tested on Firefox, Chrome, Safari, and Edge)
    • import/export of particle picks in csv file format

    For more Topaz details, visit https://github.com/tbepler/topaz/.

    For more VIA details, visit http://www.robots.ox.ac.uk/~vgg/software/via/.

     

    VIA is Copyright © 2016-2018, Abhishek Dutta,Visual Geometry Group, Oxford University and VIA Contributors.

    Copyright (c) 2016-2018, Abhishek Dutta, Visual Geometry Group, Oxford University and VIA Contributors. All rights reserved.

    Redistribution and use in source and binary forms, with or without modification, are permitted provided that the following conditions are met:

    Redistributions of source code must retain the above copyright notice, this list of conditions and the following disclaimer. Redistributions in binary form must reproduce the above copyright notice, this list of conditions and the following disclaimer in the documentation and/or other materials provided with the distribution. THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.


      Quickstart Guide

      Particle picking for training

      1) To start picking, select images or add images from absolute path.
      2) Pick 500+ particles representing as many particle orientations as possible.
          • Pick across several representative images.
          • Not all particles need to be picked from each image.
      3) Export your particle picks as a CSV file and proceed to Topaz training.

      • Save your particles and project regularly. You may load them later to continue.

      • For further help, see the Getting Started page.

      Analyzing Topaz picks

      1) Import your Topaz particle picks as a CSV file and review the picks.



      Citation: Bepler, T., Morin, A., Brasch, J., Shapiro, L., Noble, A.J., Berger, B. (2018). Positive-unlabeled convolutional neural networks for particle picking in cryo-electron micrographs. arXiv.

    Topaz Training Command Generator

    Create a command for training a network using particle picks.
    Values you may want to change are bolded. Values you need to change are underlined. Values you may want to optimize are bold-italicized.


    Fill in parameters and run the resulting command:

    Scale factor:
    Number of CPUs:
    Pixel sampling:
    Number of iterations:
    Random seed:
    Output folder:
    Input micrographs:
    topaz train --train-images /path/to/preprocessed/images/ --train-targets /path/to/training_particles.csv --k-fold 5 --fold 0 --cross-validation-seed 42 --radius 4 --model resnet8 --units 32 --dropout 0.0 --bn on --pooling none --unit-scaling 1 --ngf 32 --method GE-binomial --autoencoder 0 --pi 0.035 --l2 0 --learning-rate 0.001 --minibatch-size 256 --minibatch-balance 1/16 --epoch-size 5000 --num-epochs 10 --num-workers ## --test-batch-size 1 --device 0 --save-prefix /path/to/model --output /path/to/unique/output/model.txt

    Topaz Cross-validation Command Generator

    Create commands for cross-validating parameters for optimization. This is an optional Topaz step.
    Values you may want to change are bolded. Values you need to change are underlined. Values you may want to optimize are bold-italicized


    Fill in parameters and run the resulting command:

    Scale factor:
    Number of CPUs:
    Pixel sampling:
    Number of iterations:
    Random seed:
    Output folder:
    Input micrographs:
    topaz preprocess --scale 8 --num-workers ## --pixel-sampling 200 --niters 100 --seed 1 --verbose -o /path/to/output/preprocessed/images/ /path/to/input/images/*.mrc

    Topaz Picks Extraction Command Generator

    Create commands for extracting picks.
    Values you may want to change are bolded. Values you need to change are underlined.


    Fill in parameters and run the resulting command:

    Scale factor:
    Number of CPUs:
    Pixel sampling:
    Number of iterations:
    Random seed:
    Output folder:
    Input micrographs:
    topaz preprocess --scale 8 --num-workers ## --pixel-sampling 200 --niters 100 --seed 1 --verbose -o /path/to/output/preprocessed/images/ /path/to/input/images/*.mrc