Probably doable as a 3/4 circle positioned appropriately around the rectangular node.

Do some exploration of different styles/looks and converge on an appropriate one for the GRNsight website. Preliminary target on 2/6, prelude to completing Issue #1.

Sometimes the node gets stuck on the mouse pointer even though the mouse has unclicked. The node then gets dragged around when the mouse is moved. It seems that clicking on another node fixes the problem.

This is an issue with Firefox only it seems (Chrome is working). When loading a file, only the nodes show up with invisible (or white) edges. The mouse pointer also seems to have a charge, moving the nodes without clicking on anything.

This will probably require getting underneath the force layout algorithm.

How do we want to handle uploaded files? It seems that the normal ways to do it are to either stream the file directly or to save them as temporary files. Both methods have their own hitch that we should discuss during tomorrows meeting. I will write a preliminary implementation tonight that will likely use the first method.

Implement the file upload web service as described in the milestone. Accompany the implementation with a unit test.

We need a standalone web page that shows a sample GRN graph with a user interface to the full set of d3 force layout parameters. This way a user can interactively experiment with the settings in order to determine the best way to initially lay out a graph.

Create checkbox that determines whether the positions of the nodes are locked into place or unlocked and can be dragged by the user.
This function is part of the undo/redo sequence.

This will require some collaborative work with britainjs.

Pick an appropriate starting point from one of the d3 samples and adapt it to display the JSON representation returned by the file upload web service.

Implement the web page as described in the associated milestone. For now, it will be sufficient to display the returned JSON on the console. The next milestone will concern itself with visualization.

Let's put the website header, sidebar, and footer in centralized files so that they are easier to update.

Occasionally when the files "del_of_yap1_estimation_input.xlsx" or "Dahlquist_wt_estimation_output.xlsx", it causes a undraggable box to show in the upper right hand corner. After this, no new files can be submitted. The error doesn't happen all the time, but when it does, the message "Uncaught TypeError: Cannot read property 'getAttribute' of null" shows up in the console.

Distinct from automatic grid (Issue #25), we can just snap to grid as the user drags a node.

This is a switch on the file input plugin, should not be hard to do.

The code is ready to show some real GRNs.

In keeping with the idea of "What would Adobe Illustrator do?" we are going to implement a multi-level undo/redo function.
Key design issues:
How many levels? 5-10?
What actions can be undone/redone?
--moving positions of force graph parameter sliders
--moving positions of nodes

Some documentation that describes what makes a valid spreadsheet format would be useful.

Ask Masao for help here. Use screen to keep a long-running instance of node without having to stay logged in.

Possibly shift to Nicole if needed.

Write up documentation on how to get the GRNsight parts up and running.

Is there a specific definition of gene regulatory network I should use in the abstract?

Initial GRNsight logo, for use wherever appropriate.

We think this is mostly done; we are creating it as an issue just to do some testing and make sure.

I have a proposal for a different color for the web site that is more in the green/blue range instead of the green/yellow range: #006633

see http://www.color-hex.com/color/006633#

In that family, I also like:

005b2d and #005128

Here are some initial comments on content on the web site that needs to be added/edited.

1. On the People page, we need a header called "Principal Investigators" above the professor's names. We also need a header for "Research Assistants" below the section for Masao. Then you need to list Nicole, Britain, and Katrina like has been done on the XMLPipeDB People page.
2. On the Contact page, we need to add Dr. Fitzpatrick.
3. The Download link needs to go to the appropriate place on the GRNsight Github page.
4. On the Publications page, start a subsection entitled "Abstracts". Give the citation for the abstract that Nicole and Britain submitted to the LMU Symposium similar to what is on the XMLPipeDB Publications page. You can also include the text of the abstract there.
5. Is there a way to put the LMU logo on the upper right of the page, with the tagline "A Joint Project of the LMU Bioinformatics and Biomathematics Groups"
6. On the About page, let's be consistent with capitalization for GRNsight and GRNmap. "GRN" will always be capitalized, but "sight" and "map" will not. "Excel" does need to be capitalized. Remove "model and" from the second sentence because for now we are just doing the visualization piece.
7. On the News page, put a news item that's dated from the day of the Github site. It should say "The GRNsight Project is established on GitHub."
8. On the Documentation page, say "For current documentation, please visit the GRNsight wiki." with a link to the Running the application wiki page.
9. On the Links page, have a section called "Software" and a link to "GRNsight @ GitHub" with a link to the projct GitHub page.
   With these changes, we will have at least some minimal content on each of the web site pages.

Some items that were not already transcribed from Thursday’s whiteboard notes:

• Put self-referring arrowheads on the bottom side of the rectangle, and not the right
• Set rectangle and edge thickness to 2px
• Display border around embedded graph/artboard

Set gene names to a 12-character limit; if we detect genes that exceed this, we stop the visualization. We will need a mechanism for returning the error; web page should be set up to detect the error and notify the user appropriately.

Additional items not previously transcribed from Thursday whiteboard:

• Show current numeric value of sliders
• Place bounding box around graph element
• Experiment with slider positioning and layout
• Provide a lock/unlock slider function
• Provide a reset function (to accommodate Gravity)

After a user invokes parameter estimation on GRNmap, the resulting Excel worksheet file can then be given to the web service in order to be displayed.

We need to be more focused on the kinds of errors that can be encountered when uploading a spreadsheet. Those errors should be reported by the web service with the appropriate status codes, and the embedded web panel should relay those errors to the user.

It looks like the excel parser only works with xlsx files and not plain xls files. From what I can gather, the reason seems to be that xlsx is XML based whereas xls is not. Would it be possible to change GRNmap so that the output spreadsheet is an xlsx file instead of an xls file?

• Fix significant digits
• Fine-tune “hot region” for when the tooltip shows up (may require going beyond the title attribute)
• Provide feedback (preferably a halo) for the edge whose tooltip is visible

Follow these: https://xmlpipedb.cs.lmu.edu/biodb/fall2013/images/e/e6/PresentationGuidelines.ppt

Populate and customize the current boilerplate GRNsight information page with information about our project. The site is driven by the GitHub Pages infrastructure, so there will also be some learning involved.