Quantifying Western blots: pitfalls of densitometry

Electrophoresis. 2009 Jun;30(11):1845-55. doi: 10.1002/elps.200800720.

Abstract

Although Western blots are frequently quantified, densitometry is not documented and appears to be based merely on traditions and guesswork. Confirming previous experience, none of 100 randomly selected and systematically scanned most recent papers provided sufficient information on how Western blot results were translated into statistical values. The importance of such information, however, becomes evident from our correlations of plasma erythropoietin values of various mammals determined using RIA and Western blot densitometry. Different common densitometry procedures applied to the identical Western blot revealed p-values of these correlations ranging from 0.000013 to 0.76 reflecting the necessity of a scientifically sound basis for densitometry of Western blots. At present, the current lack of any definitions in densitometry opens the door to uncontrollable acquisition of any desired p-value. Here we provide data that define what should be considered, what avoided and what documented when quantifying Western blots.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Algorithms
  • Animals
  • Blotting, Western / methods*
  • Densitometry / methods*
  • Erythropoietin / blood*
  • Humans
  • Image Processing, Computer-Assisted / methods
  • Linear Models
  • Radioimmunoassay / methods

Substances

  • Erythropoietin