<p><strong>Background:</strong> Real-time quantitative Reverse Transcriptase Polymerase Chain Reaction (qRT-PCR) is recently used for characterization and expression analysis of miRNAs. The data from such experiments need effective analysis methods to produce reliable and high-quality data. For the miRNA prostate cancer qRT-PCR data used in this study, standard housekeeping normalization method fails due to non-stability of endogenous controls used. Therefore, identifying appropriate normalization method(s) for data analysis based on other data driven principles is an important aspect of this study.</p><p><strong>Results:</strong> In this study, different normalization methods were tested, which are available in the R packages <em>Affy</em> and <em>qpcrNorm</em> for normalization of the raw data. These methods reduce the technical variation and represent robust alternatives to the standard housekeeping normalization method. The performance of different normalization methods was evaluated statistically and compared against each other as well as with the standard housekeeping normalization method. The results suggest that <em>qpcrNorm</em> Quantile normalization method performs best for all methods tested.</p><p><strong>Conclusions:</strong> The <em>qpcrNorm</em> Quantile normalization method outperforms the other normalization methods and standard housekeeping normalization method, thus proving the hypothesis of the study. The data driven methods used in this study can be applied as standard procedures in cases where endogenous controls are not stable.</p>
Identifer | oai:union.ndltd.org:UPSALLA/oai:DiVA.org:his-4133 |
Date | January 2010 |
Creators | Deo, Ameya |
Publisher | University of Skövde, School of Life Sciences |
Source Sets | DiVA Archive at Upsalla University |
Language | English |
Detected Language | English |
Type | Student thesis, text |
Page generated in 0.0021 seconds