Help on proper/improper WB image manipulation

Use this category for questions regarding problems manipulating proteins in molecular biology applications (expression, detection, etc.)

Moderators: mdfenko, leekaming

Help on proper/improper WB image manipulation

Postby Zlikowski » Sep 28 2016 5:00 am

Hello everyone,

I apologize if this is off-topic, (admins, please delete this thread if it is) - but I have a question about western blot data representation, specifically some image manipulations. I have recently started doing westerns and I was wondering if I am doing anything that, down the road, can be considered as inappropriate or as a scientific misconduct. I have read the following:

http://www.nature.com/authors/policies/image.html
https://www.elsevier.com/editors-update/story/publishing-ethics/the-art-of-detecting-data-and-image-manipulation
http://retractionwatch.com/2012/04/03/can-we-trust-western-blots/
http://www.jci.org/articles/view/28824/version/1

and some other articles and I am still quite unsure whether my image manipulations are a big no-no. Here's an example of what I do:

1. After transfer, I dry out and scan the membrane to obtain an epi-image of the membrane and document the position of ladder/markers (https://i.imgsafe.org/b8995a5239.png)
2. I proceed with WB and obtain chemiluminescence image/detect bands on that same membrane using ImageQuant LAS 4000. (Raw image here: https://i.imgsafe.org/b8a60d96d8.png)
3. In Photoshop, I place chemo and epi membrane images on two separate layers. Then I increase the contrast of the markers on the epi image using Levels function, if necessary. Then I carefully align edges of the membrane on both layers - so that both images of the membrane, chemo+epi, are exactly one on top of the other (1:1 mapping), followed by cropping of the excess regions. Then I overlay the cropped images using the Multiply blending mode to obtain this image: https://i.imgsafe.org/b8cbc44e11.png
4. I remove the background grays on chemo image using Curves function to obtain the final image that makes me most happy: https://i.imgsafe.org/b8d619a322.png

So, am I misrepresenting data using this workflow? As far as I can tell, no data appeared or disappeared and the manipulated image in step 4. represents the original data in a correct manner. What are your thoughts on this and do you have any suggestions for a 'correct workflow', if this is unethical, wrong and or problematic. Thanks for helping out!
Zlikowski
newcomer
newcomer
 
Posts: 4
Joined: Sep 28 2016 4:49 am

Re: Help on proper/improper WB image manipulation

Postby Zlikowski » Nov 04 2016 4:49 am

Anyone?
Zlikowski
newcomer
newcomer
 
Posts: 4
Joined: Sep 28 2016 4:49 am

Re: Help on proper/improper WB image manipulation

Postby r.rosati » Nov 06 2016 12:55 pm

In my opinion, your last alteration - removing the background grey - characterizes a manipulation, since background is a variable in densitometry.
Improving the standards' contrast would practically have no big consequence, as the bands are used as a molecular weight marker and not for quantification. If for some reason you feel it's important to have a composite image, then go on, but be open: specify your process clearly. Overall though, the cleanest solution is to use the blotting image by itself, and add text indicating the MW of bands.
r.rosati
ModSquad
ModSquad
 
Posts: 2136
Joined: Nov 04 2002 10:23 am
Location: Brazil

Re: Help on proper/improper WB image manipulation

Postby Zlikowski » Nov 10 2016 6:10 am

Thanks for your input, I really appreciate it. I like to have a composite (epi+chemo) image so that anyone looking at the data can estimate themselves the sizing of the bands - I feel that this cannot hurt and that in fact it adds a tiny layer of reliability to data interpretation. As for background removal, I completely agree with you. One of the reasons I was asking about this is that, sometimes, an image of the one and the same membrane has ***less*** gray background after ***longer*** exposure under otherwise identical imager settings. I interpreted that this means that the imager itself does automatic image corrections during acquisition (Spec sheet mentions words like dark frame correction, flat frame correction, distortion correction, shading correction. Also, imager software itself has some options for image corrections following image acquisition). All of this means that I was left wondering: a) what could be considered a proper, raw image of the membrane and b) properly, legally presented data. Thanks again for your input.
Zlikowski
newcomer
newcomer
 
Posts: 4
Joined: Sep 28 2016 4:49 am

Re: Help on proper/improper WB image manipulation

Postby r.rosati » Nov 10 2016 8:35 pm

I would think journals have software to check image manipulation, think something on the lines of Fotoforensics.com. Then humans decide if any alteration is to be considered minor or not.
I agree about instruments possibly adding some aleration a priori. I know it seems unfair but this would probably pass.
r.rosati
ModSquad
ModSquad
 
Posts: 2136
Joined: Nov 04 2002 10:23 am
Location: Brazil


Return to Protein Methods

Who is online

Users browsing this forum: No registered users and 1 guest