BioSystematica
GEL IMAGERS & GEL ANALYSIS
Contact the specialists now!


Home

Contact us

  IMAGING
MiniBIS
LumiBIS

MF- ChemiBIS

Transilluminators

Printers

ANALYSIS
BioNumerics
GelCompar
TotalLab
GelQuant
Gel Analysis Service

Download PDFs
Training courses
FAQ

 

 

 

 

 

 

 

 

2009 BioSystematica

 

 

GEL IMAGING & ANALYSIS FAQ

 

1. WHAT IS THE BEST FILE FORMAT FOR GEL IMAGING AND ANALYSIS?

Gel images should be grey level (not colour) uncompressed TIFF format for analysis.

However the  JPEG format  are is excellent  for minimising size of files for storage. We do not find any detail is lost compared with TIFF for gel images. JPEG files will need to be converted back into TIFF for analysis.

2. WHAT DOES 16-BIT MEAN?

16-bit means the image is digitised into about 65K grey levels between black and white. 8-bit is only divided into 256 grey levels. The more grey levels the more accurate the detection and quantification of small bands and features by software. Certain techniques for image enhancement are also more effective with more grey levels. You will however not see much difference in images on the screen or printed.

3. HOW DO I COMPARE PATTERNS BETWEEN GELS?

When trying to match bands using software it rapidly becomes apparent that there is more distortion, both within and between gels, than was thought from just  looking at the gels.  This can be "normalised" or corrected by software so that patterns are comparable but the success of normalisation is dependent on the quality, quantity and position of reference bands present.

We frequently find when workers start using software analysis for the first time that reference (marker) bands have not been given enough attention previously.

Since distortions are only measurable by use of reference band positions that reflect the distortions actually occurring two factors must be optimised from the start of a project.

1. Reference patterns must be run in multiple lanes across the gel, ideally every 3 or 4 lanes else distortions across a gel cannot be easily normalised.

2. The bands in the reference patterns must cover the area of bands of interest in the sample patterns, ideally every 10mm or so in the area of interest.