docking-org / amis Goto Github PK

For the initial images (GPR85,GPR68) when Genotype = 0 Negative Control and Genotype = 1 Positive Control

The new sets of images are Genotype = 2, (4th category --> HCAR_Ai9_1_2)
So in this case the the Negative Control Images have "reporter-gene-cross-VEH" and the Positive Controls have "reporter-gene-cross-TAM" in their names (see 2 examples below)

Negative Control
HCAR1_Ai9_1_2_425001_f_p56_reporter-gene-cross-VEH_tongue_1723_nd_1_00001_10X_Olympus_tdTomato_e7d7cf1a1b39bf7a1f41ec8b54186506.png

Positive Control
HCAR1_Ai9_1_2_426001_f_p56_reporter-gene-cross-TAM_tongue_1723_nd_01_00001_10X_Olympus_tdTomato_045ce5fdbfa76db84b9bc76855489a62.png

(

Copied URL is not working

If an image is removed from the website (due to poor quality) , does the sequence automatically adapt?

That is 1-2-3-4-5, remove images 3 and 4 will 1-2-5 be displayed? For Example that is 00001, 00002, 00003, etc

Example:
ADGRB2_Ai9_1_2_326002_m_p56_reporter-gene-cross-TAM_eyes_19_nd_1_00001_10X_Olympus_DAPI_70b939cdc62a8f1632d8caa43c7a9a2d

Chinzo - For each set of 6 image in panel there are 4 MD5 tags - I can click and find the top 3 (on left - example below - I click on tdtomato (c4b9d1b4c183ea7122ea07b9e965e14f) and on Dapi below (14a3e39b0085cdfdc0af88405db1073c). For the bottom set I can use inspect Element on Firefox and "Search HTML" for "2133" or "2136". I just wondered if there were a simpler way.

http://amis.docking.org/img_browser?gene=GPR68&organ=heart&experiment=Ai9&sample_type=other&wavelength=tdTomato&pos_mouse_number=2130&neg_mouse_number=2133&selected_slice=1

The reason is that I could cherry-pick and create image lists for a python script to rotate images.

Thanks - Noah

Can we adjust the landing page/home page such that we can select the gene in the drop down table? i.e.:

http://amis.docking.org?gene=GPR85

would set the gene in the drop down menu to GPR85 and display the table for that gene. This will allow them to link out on specific targets to our homepage.

Thanks!

The homepage tables should pull from this table:
https://docs.google.com/spreadsheets/d/1VYAmpaQkvyN9uWsNgCovoNl6jZeIEtfv9Mz-gcfu6ec/edit?usp=sharing

They are not currently. Some syntax on the table has changed, please alert Justin English if edits to the table are necessary or if you need editing access.

"Is it possible to make rendering of double-clicked images (full size) faster? Is the bottle neck for rendering user side or server side?"

Is this hardware, software, it has something to do with local computer/internet too.

(python3_env) -sh-4.1$ source /home/justin/upload/python3_env_hidden/bin/activate
(python3_env) -sh-4.1$ python -V
Python 2.6.6
(python3_env) -sh-4.1$

" Need documentation somewhere in viewer that double clicking images with bring up full sized image."

"Can we move the tdTomato/tdTomato-RI/DAPI selector from below the prev/next buttons to beneath the image preview bar (bottom above invert color/change background color) buttons? Those buttons pertain to those images and not the larger view window above"

The + and - characters are only correct for genotypes 1 and 0 (positive and negative), as we move to different
schemes (such as genotype 2 where positive is TRAM-treated and negative is VEH-treated), the + and - will confuse molecular geneticists. So it is better to just have mice categorized under Positive or Negative Controls.

For example in : http://amis.docking.org/img_browser?gene=GPR85&organ=brain&experiment=Ai9&instrument=Histological&wavelength=tdTomato&pos_mouse_number=2505&neg_mouse_number=1537&selected_slice=14

"+ ♀ p30 2505" should now be "♀ p30 2505"

Hope that makes sense - Noah

"For empty datasets (example, negative light sheet or DAPI light sheet currently in dataset) let's display a placeholder image of black with white text N/A in center."

Check image slice padding and original image set

"User entered slice number call up for image sets"

This will be useful for image sets having hundreds or thousands of slices, scrolling though large data sets is a little slow.

To streamline user experience we should hide options that contain no data. Hide drop-down menu combinations without any data and grey out and make unselectable mouse number identifiers (buttons to the left of the images) that contain no data for the viewed gene/organ set.

See Image or contact me for Clarifications:

• Adjust png Size at bottom, rescale images to 100%,
• Move Buttons etc.
Move TdTomato, tdTomato-RI,Dapi, Invert color, Change background color, Download Tiff file, and Sharable URL
To bottom left and add some type of frame around them
• Move Navigation Buttons
From Middle to Bottom Center
• Add Controls in font
Add text for “positive controls” (top) and “negative controls” (bottom)
• Hide MD5 value tag from Web Access but not from Me
Add an Editor or Secret Mode?
• Move LUT Bar
Move from Right Side to adjacent to tdTomato RI image

please follow precedent used in ZINC. I can assist if need be.

Renames certain files by removing ".tif" string from jpg and png files (see attached file).

For example See URL = http://amis.docking.org/img_browser?gene=TAS2R4&organ=brain&experiment=Ai9&instrument=Histological&wavelength=tdTomato&pos_mouse_number=2408&neg_mouse_number=2410&selected_slice=1

rename.txt

Chinzo Running a python script to rotate 16 images on server and got this warning.

"/mnt/nfs/ex9/idg-images/python3_env/lib/python3.6/site-packages/PIL/Image.py:2735: DecompressionBombWarning: Image size (173067180 pixels) exceeds limit of 89478485 pix els, could be decompression bomb DOS attack.

DecompressionBombWarning,"

See attached file.

"When switching to cleared the organ dropdown menu does not update and shows all histology available organs. When switching from cleared after browsing the images the organ dropdown list does not show all available organs. Can we push an auto refresh to this dropdown when changing Sample type?"

"Make black background by default and switch to white white change background on the image browser? Think this looks really good."

"Light sheet image display is deformed. Do not rescale unevenly for any datasets. Maintain image size ratio."

I replaced some images this morning, MTNRIA "Jej", "Liver" and "Colon" with renamed images to "jej", "liver", and "colon". The
misnamed images are no longer there, but the categories seem to be there

For Example:
http://amis.docking.org/img_browser?gene=MTNRIA&organ=Jej&instrument=Histological&wavelength=tdTomato&pos_mouse_number=326003&selected_slice=1

Can we get access to Google Analytics or other webserver user analysis? Useful for Justin to push usership at IDG etc. meetings. Also for us to see how people are using it and determine where data enrichment would be valued.

Chinzo - Is it possible to run Sciimage Python scripts on server. Would I need to load libraries locally or is there an approach (i.e. ananconda or pip) already in place.

Thanks -Noah

Chinzo - I'd like the ability to run a script or scripts that will remove images (all) or image list from database. And Update the database to contain the tifs currently on the image server.

Is that possible? - Noah

Would it be possible to only display range of slices images around selected Slice?

That is if there are 900 slices - updating all of the pngs takes a long time and,

Explicit choice of slice #: would it be possible when the users selects image 112 - that the Slices around #112 are shown.

Scrolling Through Slices: Would there be a way to wait a certain amount of time until
an image Slice position (#) is selected and then populate the range of Slice Images around
that choice?

For the expression summary table tdTomato images on the homepage please swap display of TIF thumbnails associated with the listed MD5 for RI images of the same MD5 name. DAPI images are fine and do not need to be changed.

"Shopping cart or image series download capability so users can pull multiple images in batch. Users should only download TIF files. Thoughts on how to integrate this?"