Protein electrophoresis algorithm

  • Thread starter Thread starter victorqed
  • Start date Start date
  • Tags Tags
    Algorithm Protein
AI Thread Summary
The discussion revolves around the challenges faced in developing an application to analyze electrophoresis gels and accurately quantify protein levels, specifically albumin and other fractions. The user reports discrepancies between their application's results and the expected values from a standard solution, with albumin levels appearing significantly lower and other proteins higher than anticipated. They have tested their application against existing software like ImageJ, which yielded similar incorrect results. Key concerns include the potential for detector saturation affecting albumin intensity measurements and the suitability of the scanner used for capturing gel images. Suggestions include adjusting the scanner settings, diluting samples, and ensuring proper lighting conditions to enhance measurement accuracy. The user acknowledges the need for a more sensitive scanner and is exploring solutions to improve the software's ability to adjust protein peak intensities. The discussion highlights the importance of accurate calibration and the impact of equipment quality on electrophoresis analysis.
victorqed
Messages
8
Reaction score
0
Hello

I'm trying to implement a application that will scan the electrophoresis gels and draw a graphic of the proteins: albumin, alpha1, alpha2, beta1, beta2 and gamma.
The problem is that the resulted graphic is not as it should be. Albumin percentage is too low and the rest of the proteins are too high. Also, looking at the my graphic I can see that albumin is very low.

I tried to make a comparative test with many applications that already exist on the web and all of them are giving me the same values as I calculated. But this values are wrong so it is clear that I'm missing something essential.

One of this applications is ImageJ(open source).
http://imagej.nih.gov/ij/
http://lukemiller.org/index.php/2010/11/analyzing-gels-and-western-blots-with-image-j/

Example:
I have attached a scan of a normal control done on a Sebia gel.
Correct values are:
Albumin : 66,05%
Alpha1 : 2,47%
Alpha2 : 9,14%
Beta1 : 6,9%
Beta2 : 4,33%
Gamma : 11,11%

Values obtained with my app and ImageJ are:
Albumin: 46,76%
Alpha1 : 3,77%
Alpha2 : 15,05%
Beta1 : 11,78%
Beta2 : 5,36%
Gamma : 17,29%

How can I reproduce the correct values?
Thank you
 

Attachments

  • CN.jpg
    CN.jpg
    1.8 KB · Views: 559
Last edited:
Biology news on Phys.org
What values are you reporting? Migration distances? Calculated molecular weights? Intensities?
 
Values are in percents and are calculated as part of the total area of the resulting graphic.
I will edit the original post to add % sign.
 
So, you are measuring the intensity of each band in the gel, then dividing by the total intensity. How do you know your correct values are correct? Are you making up a standard solution with known concentrations of each protein? Is there a way to check whether you have made this standard solution correctly?

Another possibility is that the intensity from the albumin (the top band I presume), is saturating your detector, so your measurement is lower than the amount of protein present. Diluting the sample and running again may help (or changing the settings on your detector). How are you quantifying the intensity of each band? What type of staining are you using to visualize the proteins?
 
  • Like
Likes 1 person
The correct values are from a prepared solution, made by a specialized producer.
Solution has these concentrations of proteins from fabrication.
So the correct values are 100% correct.

You are right, the scanner is not specially made for electrophoresis and it might not be sensitive enough to see the peak of the Albumin.
I will try to fix the problem software by increasing the Albumin peak until I reach the target values.

If Albumin intensity is the problem than I should reach all protein values at the same time as I increase albumin.

Thank you :)
 
So, the scanner is the problem.
There is no complete software solution, I need a scanner that will see all the Albumin.
Where can I find one?
 
Have you eliminated the possibility of differential dye binding? Also if you aren't using a light box type of instrument, then you have all sorts of things to worry about. For example the lighting varying from place to place on the gel, as well as the inherent differences in the pixel sensitivities of the camera (or lens vignetting, dust etc). Some of these camera problems can be fixed easily, others not so much. You can google for something like flat-frame calibration to see what can be done with regard to the differences in the inherent pixel sensitivities.
 
  • Like
Likes 1 person
All solutions are standard, differential dye binding shouldn't be a problem but I will ask.
Light could be a new problem. :)
 

Hot Threads

Recent Insights

Back
Top