Relationship between detection limit and bias of accuracy of quantification of RNA by RT-PCR

Nested PCR bias: a case study of Pseudomonas spp. in soil microcosms

Nested PCR methods combined with denaturing gradient gel electrophoresis (DGGE) are widely used for the detection of low copy number species or for the analysis of group-specific community profiles. With an appropriate number of PCR cycles during the first round of amplification, initial differences in the copy numbers of different DNA fragments that are targeted can be maintained during the second round without significant bias. However, if an excessive number of cycles in used in the first round, relative differences in the copy numbers of the targeted sequences can be obscured. Here we demonstrate the effect of "nested PCR bias" in a case study with PCR-DGGE of 16S rRNA gene sequences targeting Pseudomonas spp. following exposure of soil to naphthalene vapors. Our results demonstrate artifacts caused by nested PCR bias can be substantially minimized by calibrating the number of first round PCR cycles, thereby preserving the ability to obtain semiquantitative data for evaluating changes in gene copy numbers over time.

Distribution-insensitive cluster analysis in SAS on real-time PCR gene expression data of steadily expressed genes

Cluster analysis is a tool often employed in the micro-array techniques but used less in the real-time PCR. Herein we present core SAS code that instead of the Euclidian distances takes correlation coefficient as a dissimilarity measure. The dissimilarity measure is made robust using a rank-order correlation coefficient rather than a parametric one. There is no need for an overall probability adjustment like in scoring methods based on repeated pair-wise comparisons. The rank-order correlation matrix gives a good base for the clustering procedure of gene expression data obtained by real-time RT-PCR as it disregards the different expression levels. Associated with each cluster is a linear combination of the variables in the cluster, which is the first principal component. Large set of variables can then be replaced by the set of cluster components with little loss of information. In this way, distinct clusters containing unregulated housekeeping genes along with other steadily expressed genes can be disclosed and utilized for standardization purposes. Simulated data in parallel with the data from a biological experiment were taken to validate the SAS macro. For both cases, good intuitive results were obtained.