Analysis of similarity of electrophoretic patterns in mRNA differential display

Abstract

We present a novel method of statistical analysis for the comparison of electrophoretic data. The method is based on the squared Euclidian distance of normalized signal data vectors of electrophoretic lanes. The differences in the electrophoretic patterns are evaluated by a statistical test based on Hubert's statistics which measures the significance of the signal grouping. We demonstrate the validity and applicability of the method in a large data set derived from automated fluorescent mRNA differential display analysis of the expression of acute-phase proteins during experimental Escherichia coli infection in mice. The current testing method is capable of finding theoretically similar natural groupings to be similar in a statistically significant way whereas theoretically dissimilar or random groupings can be recognized to be artifactual. We also show how the calculated pairwise signal distances can be utilized in methodological problem solving. These analytical methods can be applied to the study of other related problems of similarity analysis of electrophoretic patterns, and also provide useful tools for the development of automated recognition of differentially expressed mRNAs.

Automated detection of differentially expressed fragments in mRNA differential display

“…All differentially displayed fragments in these highly similar samples within a group are assumed to be linked to the experimental infection. We have shown earlier that lane data within these theoretically similar groups show statistical cohesion [6]. In the current study, there were 624 groups of three lanes producing 1872 lane pairs to be analyzed on the fragment-by-fragment basis with respect to the presence or absence of differentially displayed fragments.…”

“…The requirement of similar background is considered by dividing the distance by the average difference of the environment of other L* -1 matched peak pairs. Here, the total distance D* is used in the place of the Euclidian distance, previously employed in the analysis of electrophoretic data on a lane-by-lane basis [6]. In statistical terms, the observed value of S k is used to test whether the distance of p* k is an outlier in p*.…”

“…We have previosly shown that numerical mRNA differential display data can be used to analyze statistically the relatedness of electropherograms [6]. Here we present an algorithm that performs a more detailed analysis of lanes on a fragment-by-fragment basis.…”

mRNA differential display of gene expression in colonic carcinoma

“…If the mispriming and the rest of the process can be rendered reproducible enough, even small changes in mRNA expression may be detected reliably. We have shown previously by the similarity analysis of numerical DD-PCR data [20] that this goal can probably be achieved.…”

“…Allelic frequency can be confirmed by analyzing the corresponding fragments from the denaturing DD-PCR-gel. It is evident that this kind of approach is best solved by automated analysis of numerical DD-PCR data [20,21] produced in an automated multicapillary sequencer. A modern 384-capillary multicolor electrophoretic device can analyze several million DD-PCR fragments in 24 h. The 2880 electrophoretic samples of the current study could be analyzed in 3 h.…”