General

This page is to help you learn more about how to use Mozak and solve common problems when using the app.

About

Thank you so much for trying Mozak. Our current goal is to be able to give Mozak players (non-experts) the tools needed to construct expert-level neuron reconstructions.

Technical requirements

  • We currently support all recent versions of Chrome, and Firefox versions 47 and higher. The most current Safari on Mac should also function as well as the latest Microsoft Edge.
  • It's highly recommended, but not required, to play with a mouse and laptop/desktop. We don't think the game will work well on a touchpad, but we'd like to hear about your experiences if you try it.
  • Mobile and touch devices are supported, although there are issues on iOS we hope to resolve in the future.

Hints for common startup problems

  • Created account, but unable to log in

    Check the email you registered your account with. You should have received an activation link. Click the activation link to activate your account and log in.

  • The game loses my place when I go to another part of the site or close and reopen the page.

    Make sure you have cookies enabled. Progress is only saved across machines if you create an account.

  • What if nothing shows up on screen?

    Try these solutions first:

    • Try clearing your browser cache. Sometimes when we update the websites, certain files get out of date and cause errors unless a fresh copy is fetched. Instructions to clear the cache for a specific site on Chrome and Firefox. can be found here. For Safari on Mac, Go to Safari->Preferences->Privacy->Manage Website Data. Search for mozak and ‘Remove’ the data for the site.
    • Certain browser extensions (like popup blockers) or anti-virus software may cause errors when trying to run the application. If you have such programs installed and are encountering problems with the site, consider adding a security exception to those programs.

  • I went through those steps and am still having problems.

    Try opening up the browsers development console. In Chrome and Firefox, this is with Ctrl+Shift+I. On Safari you can go to 'Develop'->'Show Javascript Console' This should bring up a new window with several tabs. Click on the one that says 'Console'. The console shows warnings and errors that occurred in the application. By default the most serious errors are in red, if possible copy the output and paste it into a forum thread so the developers can take a look.

How Mozak Works

Each neuron in Mozak is a single, extremely large, image — so large that your computer would never be able to load the whole thing at once. Mozak cuts it down to size by allowing you to see small parts of the image in full detail, but also letting you view less-detailed versions of the image from farther away, so you can get a better idea what's going on on a larger scale.

Map View

The Map View provides an easy way to examine different parts of the neuron quickly, albeit only in two dimensions. While moving around, a transparent green box will indicate the region that you load when going into the Edit view. When you have found a portion of the neuron image needing annotation, click the Toggle 2D to 3D button (a compass icon) to load a portion of the neuron into the three-dimensional annotation view. Note that in this view mode you cannot make edits to the traces.

These are the controls to move the camera while in this view:

  • Panning: Click and drag with the Right Mouse Button to slide different parts of the neuron into view. On touch devices drag with one finger.
  • Zooming: Use the mouse wheel to zoom closer and further away from the neuron. On touch devices, use two fingers and perform a pinch or spread gesture.

Annotation/Edit View

The Annotation or Edit View is where you'll be able to use the various tracing tools, described below. If you want to move to a different portion of the neuron that is far away from your current view, you can click the Toggle 2D to 3D button again to return to map view. To look at a different section of the image without leaving the edit mode, you can click on existing nodes to force the application to load the area around that node. There are also features described further below that will allow you to expand the amount of 3D volume that can be seen at once.

  1. Settings Menu - Mainly for advanced uses, this menu contains settings to adjust tool behavior, 3D volume appearance, and trace appearance. Also allows for sign out/log in.
  2. Status - General status indicator area. When using some of the tracing tools, this shows progress of the tool or whether it is in a usable state.
  3. Score Tab - Shows amount of 'work' you have contributed to the current neuron and how much as been incorporated into consensus.
  4. Tags - Tags can be placed on nodes by pressing the 'T' key. Use this button search for tags you have added on the current challenge.
  5. 3D Image Clipping
  6. Chat Tab - Expand to view public chat channel
  7. 3D Volume and tracing area - This is where the neuron image and reconstruction will appear and where you will make edits. See here to control 3D volume loading.
  8. Quests Tab - Shows current quests that you have not yet completed. Click or hover over the icon to view details about the quest.
  9. Reconstruction Tools Dock
  10. Brightness/Contrast - Click the invert icon at the top of the slider to invert the colors of the neuron image. Use the slider adjust the contrast of the image, making it appear lighter or darker. This may help you identify difficult-to-see sections.
  11. Features/Settings Status Bar
  12. Neuron Picker Tab - Expand this tab to switch to different neuron if multiple are active at one time.
  13. Legend Information - Shows information camera controls and additional controls for the current tool you are one as quick reference.
  14. Viewpoint Url - While in 3D mode, click to copy a url at the exact viewpoint you are currently at. You can use this url to report bugs or share a particular location of the challenge you have a question about in the forum.

Camera Controls in Annotation/Edit View

These are the general controls to move the camera while in this view:

  • Zoom In/Out: Use your mouse's scroll wheel. On touch devices use two fingers and perform a pinch or spread gesture.
  • Panning/Shifting View: Hold down Shift while dragging the Right mouse button to pan the 3D volume. You can also use the keyboard arrow keys. On touch devices, place three fingers on the screen and drag them to pan.
  • Rotate Camera: To view the neuron from a different perspective, hold and drag with your right mouse button. This is helpful when you feel like you've completed a section and want to view it from another angle, or when you want to get a better look at details that are hidden behind other parts of the neuron. If you are using a trackpad that does not have any left and right click buttons, then you can perform the rotate by using a two finger press and then drag. On touch devices, place two fingers on the screen and drag them to rotate.

Clipping the 3D image

Clipping the current 3D image volume can make it easier to follow specific signal paths in areas that are dense in signal or have image noise. They also can help improve the performance and quality of the Virtual Finger and Magic Tracer tracing tools by reducing the amount of data they have the process.

The Mozak clipping tools allow you to control the axis from which layers of the image are clipped from the current view and the amount of layers. The axis, which determines the direction from which layers are clipped, are controlled by the buttons on the right side of the slider. While the slider itself controls the amount clipped. The smaller the sliding window, the smaller the number of layers of the image will be visible.

By default clipping will be set along the z-axis of the image, which will hide layers going toward you and away from you. Clipping along the x-axis hides layers going from left to right and the y-axis from top to bottom. These directions are unaffected by orientation of the camera. There is also a special camera direction axis, which clips layers on the direction the camera is oriented in. The orientation can be locked so it doesn't change when you rotate the camera.

Hold down 'Q' and use the Mouse Wheel to change the size of the clipping window. Hold down 'W' and use the Mouse Wheel to shift the clipping window. See other shortcuts in the '3D Volume Loading + View' section of the keybinding menu by pressing 'K' while in the app.

Status Bar Features

The status bar provides quick access to important features and settings that can be adjusted while tracing.

  1. Resolution Level
  2. Brick Loading Settings
  3. Brick Auto-load
  4. Auto Recenter
  5. Focus Sphere View
  6. Trace View
  7. Trace Color Settings
  8. Node Labeling/Typing
  9. Help Page
  10. Keybindings
  11. Update: Button will appear if there are updates to the app that require downloading

Back to top

3d Volume and Loading

The 3D image volume is where you will be adding and editing the trace reconstruction for a neuron cell. This image data is too large to be downloaded and viewed all at once, as data size can range from several Gigabytes in total for smaller cell samples and up to hundreds of Gb for larger ones. As such, there are a number of features and settings in Mozak to control the location and how much of the image is downloaded and displayed to you.

Resolution Level

There are usually multiple resolution levels you can view a 3D image in. The higher the resolution level, the more detailed and zoom in representation of the image. However, they also require more data to be downloaded and a smaller overall area of the cell can be viewed at once compared to a lower resolution. If want you to be able to see more of the 3D volume at once, consider switching to a lower resolution as sometimes for a particular area the decrease in image quality may not noticeably affect your ability to identify and trace signal.

The current resolution level can be viewed at the bottom left of the screen. Hover over or tap on the indicator to bring up a menu to lower or increase the current resolution by one.

Bricks

In Mozak, the 3D volume is divided into units called bricks. A brick is simply a box made up of pixels and is the smallest unit of 3D data that be be downloaded and displayed. Based on the specifications of your device (mainly RAM and Graphics Card), there is a limit to the number of bricks that can be displayed at once.

Controlling What/When to load

There are two shapes loaded bricks can appear in. You can toggle shape via the indicator button directly to the right of the resolution level indicator and at the bottom of the screen. Press and hold the button to bring up a settings menu.

  1. Box: The 3D volume will appear as a contiguous rectangle centered on the currently selected node. You can adjust the size of the box along each of the X, Y, and Z dimensions. (Warning: Depending on your computer, setting this to too large a volume may crash the app. See here to fix.

  2. Around Traces: Bricks will load wrapped around existing traces starting at the currently selected node.

    You can adjust the following attributes in this mode:

    • Adjacent Brick Radius – Determines the ‘thickness’ of the 3D volume that wraps around the traces. Increasing will allow you to see more of the surrounding area around a trace but at the expense of seeing less traces covered.
    • Max Total Bricks – Toggle the slider to view more/less total bricks. (Warning: Depending on your computer, setting this to too large a value may crash the app. See here to fix )
    • Look Ahead Voxel Distance – When you are extending traces, you’ll often want to see bricks ‘ahead’ of where you currently are. This is the distance ahead of the currently selected node to render 3D bricks, by increasing it more of the volume will be rendered in front of the selected node.

You can also control when new bricks load. In the automatic loading mode, new bricks are loaded and displayed whenever you create a new trace or when you select a node.

In manual mode you must use the keyboard hotkey ‘R’ to load new bricks. Downloading and rendering new bricks can be slow and interrupt your tracing, so turning off the auto-loading is useful if you want to focus on making several edits in a fixed area that is dense with signal.

How to Play: Tracing Tools

The neuron reconstructions you build in Mozak are made of a skeleton of nodes connected by lines. You can also see the reconstructions that other players have made. These contributions are part of something called consensus, which is the common work that multiple people have contributed to. Below are the tools available in Mozak that you will use to create and make edits to the reconstruction.

Reconstruction Tool Dock

  1. Virtual Finger
  2. Connect-the-Dots (Single Slice)
  3. Connect-the-Dots 3D
  4. Undo
  5. Redo
  6. Magic Tracer
  7. Delete Node
  8. Delete Branch
  9. Approve Trace
  10. Map Mode to Edit mode

Virtual Finger

In Virtual Finger mode, you create traces by holding down the left mouse button and dragging the cursor along the path of the signal. A bluish preview path will appear where you dragged. After you release the mouse, the application will calculate and render the appropriate final trace segment.

If you move your mouse near an existing node it will highlight. You can click on a highlighted node to select it. A selected node will act as a starting point for your next trace path, any new path you create will be connected to that selected node. When you create new nodes in a path, the last node in that path will automatically be selected.

After you start tracing, if you move your mouse near a part of your reconstruction, you can highlight an existing node to connect to it. When you release the mouse the path created by Virtual Finger will connect to that end point.

Connect-the-dots mode

Some parts of the neuron are too dim to trace with Virtual Finger. When you select the connect-the-dots tool, you shift from a 3D view of the neuron to seeing a single 2D z-slice through the neuron, somewhere between its front and its back. In this view, an individual neuron branch looks like a series of bright spots and lines on neighboring z-slices.

You can move between slices by rolling your mouse wheel up and down. When you see one of these bright spots, you can place a node on it by left-clicking, and build up a branch by following the branch between planes and dropping nodes in a line as you go. Nodes that you immediately place will appear a different color, after a short delay or by right-clicking they will be saved as part of your reconstruction.

You can also use the connect-the-dots tool to connect two branches. But be careful to not make a loop! Sometimes its hard to tell if connecting to an existing node will make a loop. Mozak will warn you with a message if it would. You'll also see any potential loops highlighted in bright red.

In some situations with very close branches you'll need to place a node in a location that triggers the loop detection. If you're very certain that the node is in the right place, you can over-ride loop detection by turning Snap to Node off in the Settings Menu. Mozak will still warn you about potential loops, but you'll be able to place nodes. Remember! Neurons don't have loops, so neither should your reconstruction.

  • Mouse Wheel: Change z-slice
  • Hold down '<' + Mouse Wheel: Zoom in and out
  • Left Click: Place a node
  • The currently selected node will be highlighted orange. Any new nodes placed will connect to it. Left clicking on it will unselect it.

Connect-the-dots 3D mode

In this mode you will have a 3D view of the volume but will place individual nodes one at a time. Hover over the signal where you want to add a node, the application will place a preview sphere over the brightest pixel it has found near the cursor. Left click again to confirm placing a node at that point. Like with the single slice Connect-the-dots, there is a short delay before the nodes that you just placed will be saved, after which they change colors. Note that there might be a small delay before it can find a pixel or it might not be able to find a pixel at all if the signal is too faint.

Deleting Nodes, Edges, and Branches

Sometimes you'll make mistakes, but you can fix them. For small mistakes, select the Delete Node tool (with the eraser icon), then left-click to delete a single node. You can also use this to delete individual edges, just mouse over and click just like a node. On touch devices, drag one finger over the portion you want to delete. After it becomes highlighted, release you finger to delete.

For bigger mistakes, you can delete an entire continuous path. Select the Delete Branch tool (with the trash icon), then left-click on the branch to delete. On touch devices, perform the same action when using the delete node/edge tool. Drag one finger over the portion you want to delete to highlight it and release your finger to delete it.

Magic Tracer

The Magic Tracer tool traces a signal path with just one click. Just like in Connect-the-dots 3D you first need to hover over a starting point of the path you wish to create. A preview sphere will appear, left click to confirm that as the starting point of the generated path. After a short delay you should start seeing a preview path being drawn. It might take several seconds for final path to be created. When the calculation is finished, the preview path will disappear.

Placing Single Node

If you ever want to place a single new node or create a connection without having to switch to one of the Connect-the-dots tools, simply hold down shift and left click. This will place nodes just as if you were in the Connect-the-dots 3D mode. Releasing the shift key will commit the changes.

Approving Other User Traces

After a sufficient amount of work has been done on a neuron, the ability to see other users' traces that are not yet incorporated into consensus may become available. By tracing in these areas you are likely to form consensus. Eventually you will also unlock a tool mode that will allow you to click on another user trace to approve it, which creates a copy of that segment belonging to you. Hover over and click on the faint traces to approve them if they look correct.

Undo and Redo

The undo action will revert the most recent trace edit actions (i.e. you added or removed nodes from the reconstruction). While redo will reapply actions that you most recently performed an undo on.

Map Mode to Edit mode

The button allows for switching back and forth between map mode and edit mode. Map mode is used for faster navigation to different areas in the neuron (it is usually 2D image). Edit mode is the 3D image volume where you add and delete traces on the reconstruction for the neuron. Commonly you want to switch to map mode if you want view a different area of the neuron that is both far from your current view and does not yet have existing traces from which you can select to load the 3D bricks around.

How to Play: Tips and Advanced Controls

Here are tips and controls you can use to make tracing more accurate and efficient.

Selecting Nodes

To extend and grow the existing reconstruction you will need to select a node as a starting point. Hover over and click on a node to select it. If you need to trace in a different area, you can use the 'S' key or the button in the bottom status bar to quickly unselect any currently selected node.

Quickly Deleting Nodes

While using a trace tool, you can quickly delete a trace by holding down the 'A' key. If a node is selected, hover over another node and click to delete the entire path between the selected node and the one you hovered over. If no node is selected then hover over and click on a node or edge to remove just that element. Simply release the key to go back to tracing. This saves you time from having to switch to one of the delete tools and then back to a trace tool.

Image Clipping when Tracing

Remember to use the image clipping slider if the current 3D volume is dense with signal or noise. This may help improve the quality of paths created using Magic Tracer or Virtual Finger.

Accidentally Selecting a Node

Sometimes when making edits to an area with existing traces already, clicking in that area unintentionally selects a node instead of placing a new one. To disable node selection, press and hold the node selection option in the bottom status to bring up the settings menu. You can toggle on/off whether node selection is enabled at all. Remember to re-enable it once you are done tracing in that area.

Back to top

Keyboard Shortcuts

Press the 'K' key to see the entire list of keybindings. These are some of the most commonly used ones when you are starting out.

  • V : Switch to Virtual Finger tracing mode
  • C : Switch to Connect-the-Dots tracing mode
  • Shift+C : Switch to Connect-the-Dots 3D tracing mode
  • Shift+D : Switch to Delete Single Node mode
  • D : Switch to Delete Branch mode
  • Ctrl+Z : Undo trace or delete action
  • Ctrl+Shift+Z : Redo trace or delete action
  • Shift+V : Switch to Magic Tracer mode
  • A : Toggle auto z-adjustment for Connect-the-Dots tracing
  • S : Unselect Currently Selected Node
  • Arrows ←↑↓→ (all computers) Translate current view
  • Shift and Right Mouse Button click/hold and drag: Another way to translate view(moves view in direction of mouse drag)
  • Spacebar : Re-centers the view along X, Y and Z axis, and focuses on the front of the challenge
  • Mouse Wheel : Zoom in and out the view
  • Google Chrome and Mozilla Firefox: Ctrl and + / Ctrl and -: Increases / decreases the scale of the app

Back to top

How to Play: Common Errors/Mistakes

Z-Jumps

Z-jumps happen when Virtual Finger gets confused about what z-slice a branch is on. It creates a segment that inappropriately stretches between z-slices that are great distances away. You can use the delete tool and delete single nodes to fix this issue.

Tracing False Positive

Sometimes you will encounter strands that look like a part of the neuron but are not really. These are false positives and should be avoided as the scientists need to cleanup/remove them during their evaluation. As a rule of thumb, if the strand you are looking at does not eventually connect back to the main neuron structure you should avoid tracing it.

Comment area looks empty

If the bounding box marked by a comment looks like it's empty, try using Connect-the-Dots mode to look through each z-slice for things you can't see in 3D. If you still can't see anything, try inverting colors or dragging the contrast slider to get a better look. Use the H key to turn off consensus traces if those are interfering.

Back to top

FAQ

What is consensus?

Consensus is an on-going process that looks at all the traces that users have contributed to a particular neuron and merges together branches where multiple people have traced the same branch. This results in a single output reconstruction that eventually gets sent to the scientists at AIBS for analysis. Any particular neuron might have several consensus runs applied to it during the time it is open.

Why don't I have a score yet?

Scores are calculated after a consensus run has finished. Depending on the number of people that have traced on a neuron, it can take several hours or even a few days for it to complete.

What do the colors mean?

The colors of the traces you see will differ depending on what view mode you have turned on. You can access and change the view mode and customize the colors in the status bar at the bottom of the screen.

If you have the 'Trace View' turned on:

Your traces

Consensus traces

If you have the 'Connect View' turned on:

A neurite that is already connected back to the soma.

A neurite that still needs to be connected back to the soma.

What does the option 'Z-Clip in CtD' do?

This option clips the traces that are not close to the current z-slice you are viewing when you are in Connect-the-Dots mode. It does the same if in Virtual Finger mode and you have adjusted the viewable volume using the Mouse Wheel while either the Shift or Alt key is held down. It can prove to be a very handy tool option to remove extra visual noise when identifying faint branches. One important note is that the 'Thick Lines' option must also be turned on for it to work. The downside is that 'Thick Lines' might reduce application performance.

The browser keeps crashing

You might be trying to load too many bricks at once, which will cause an out of memory error. To fix this, reload the page and then quickly press ‘Shift + R’. This resets the brick loading settings to defaults.

The application feels sluggish in general.

There are multiple factors that can affect the slowness of the Mozak application.

Downloading and rendering new 3D image bricks can be an expensive operation for your computer. By turning off Auto-Load bricks feature, new bricks will only be downloaded when you manually press the ‘R’ key.

Similarly, try reducing the amount of bricks being shown to a smaller value.

IMPORTANT!

The browser you are currently using may not be compatible. Please use the most recent Chrome or Firefox for the best experience.