You cold try this free software http://www.licor.com/bio/educational_resources/webinars_videos/player.jsp?id=108

Using ImageJ you can filter out the noise. You don't need the signal to be 3× the background.

When acquiring the image, try to obtain the highest contrast possible with your image acquisition software. Pre-process your image with linear filters to maintain your band signal and then proceed to do a densitometry using the ImageJ plugin.

For more detailed assistance, you need to share an example of your particular scan.

ImageJ suggestions on image acquisition for densitometry:

http://imagej.nih.gov/ij/docs/guide/146-30.html#infobox:Densitometry

Other densitometry with ImageJ tutorials:

http://lukemiller.org/index.php/2010/11/analyzing-gels-and-western-blots-with-image-j/

http://howtowesternblot.net/data-analysis-3/quantification/

And a video tutorial for the reading-challenged ;-)

http://imagejdocu.tudor.lu/doku.php?id=video:analysis:gel_quantification_analysis

http://imagej.nih.gov/ij/docs/guide/146-30.html#infobox:Densitometry

If the above suggestions don't work with the image you have right now, you could try to increase the actual signal in your samples before imaging:

-- Load more protein (I've loaded up to 80ug per well on a standard SDS-PAGE gel, and I know people who load even 100ug).

-- Use a more sensitive ECL kit, such as the Pierce SuperSignal West Femto http://www.piercenet.com/browse.cfm?fldID=01041106

Other pre-processing (pre-gel) options:

-- Primary antibody optimization as Matthew suggests above.

-- If you have low SNR due to high background, try adding 0.1% Tween to your primary antibody.

-- Likewise, if you have high background, try adding 0.1% Tween and/or 0.01% SDS to your secondary antibody.

-- If you have access to an infrared scanner, fluorescence has a larger linear range than chemiluminescence so you might be able to increase your SNR by using a fluorophore-conjugated secondary antibody.

Thank you all for your replies! I am so grateful for your input and hope to be able to pay it forward soon!

To Rafael:

thanks for the fantastic list of resources. i've been using luke miller's tutorials.

- what do you mean linear filter? is this something on the imaging system or an option on ImageJ? on the UVP system, there are pre-made settings for chemi-blots (i'm not sure what the associated filter settings are - is this what i should look into?). right now i expose the membranes for 25 minutes (which is the longest exposure time possible) so i'm assuming the contrast is as high as possible.

- the reason i need 3-fold SNR is because that is the confidence threshold my lab suggested. is there any reason to believe that data below this threshold is still useable?

- i've attached a sample blot. i can't tell you too many details other than the middle row is GAPDH and the top and bottom rows are proteins of interest. if you can suggest specific tools/protocol on imageJ that would help analyze this quality of western, that would be fantastic.

To Irit:

- thanks for your suggestions!

- i'm currently using the M-PER lysis buffer to isolate enough of my protein of interest from nuclear/ER membrane. as a result, the protein concentration is such that the maximum i can load is about 40ug (because the max volume allowed in gel is ~40ul). do you have any suggestions of better lysis buffers for membrane embedded proteins so that i have a more concentrated sample/will be able to load more protein into an SDS-PAGE?

- or alternatively, are there gels that will let me load a larger volume?

- i will look into the femto imaging substrates kit - thank you!

To Traimate:

- thanks for your post!

- i am in the processing of trying the free software you suggested.

To Matthew:

- thank you for your advice! i will definitely try blocking for a longer period of time and adding to the wash time.

- i'm not sure what sequential image integration is, but i will definitely look into it before i try rolling ball background subtraction.

To Girish:

- thanks for your input!

Thank you all once again for your input. I really appreciate it :)