Criteria for Candida albicans numerical analysis based on electrophoretic protein patterns

ABSTRACT

Even in normalized gels, some problems with the choice and position of protein bands always have hindered the processing of electrophoretic data. We have developed a way to establish which criterion best fits the necessities in order to maximize the similarity indexes for numerical analysis. Some repetitions of a Candida albicans strain were carried out in eleven different gels. After staining, the bands were scored in numbers within ranges of +/- n values with increases of one quarter steps (+/- 0.25 kDa, +/- 0.5 kDa, +/- 0.75 kDa, +/- 1.0 kDa, +/- 1.25 kDa) that will limit the possibility of variation for the same bands that will appear on the repetitions. Using this criterion, we have determined that values scored within +/- 1.25 kDa could optimize the minimum limiting value of similarity in different dendrograms built.