Local mean normalization of microarray element signal intensities across an array surface: quality control and...

Area of Science:

• Bioinformatics
• Genomics
• Molecular Biology

Background:

• Microarray data can contain spatially systematic artifacts introduced during printing, hybridization, washing, or imaging.
• These artifacts can compromise the accuracy of gene expression analysis.
• Existing methods may not adequately address spatially dependent variations in signal intensity.

Purpose of the Study:

• To present a novel methodology for normalizing element signal intensities on DNA microarrays.
• To enable the detection and correction of spatially systematic artifacts in microarray data.
• To provide a user-friendly tool for quality control and data correction in gene expression analysis.

Main Methods:

• Local mean normalization using element signal intensities calculated across the microarray surface.
• Algorithms designed to correct artifacts that vary spatially across the array.
• Development of the methodology in the R statistical language with a web implementation.

Main Results:

• Demonstrated correction of spatially systematic artifacts using only array element signal intensities.
• Validation of the local mean normalization process for quality control and data correction.
• Successful implementation as an interactive, user-friendly web tool.

Conclusions:

• Local mean normalization is an effective method for correcting spatial artifacts in DNA microarray data.
• The freely available web tool facilitates accessible use of this normalization technique for researchers.
• This approach improves the reliability and accuracy of gene expression data analysis.