I am doing survival analysis. separating the patients in groups of low and high exression of protein A gave no significant difference in survival analysis.
However, by studying correlation between exact expression % of the protein and exact survival time of the patients, there was significant correlation (spearman). What does this mean?
I must repeatedly and sadly say that distributions of the natural phenomena are obeyed by non-parametric statistics, and are not obeyed by parametric statistics such as Gauss distribution (Normal distribution), Binomial distribution (Bernoulli distribution), or Poisson distribution.
We have used non-parametric statistics in the study of IDDM urines (please see file; Dr. Terentyeva Urine BIN).
By the way, quantitative protein-expression study should be performed by PDMD (Protein-Direct-Microsequencing-Deciphering) method (please see file; HepG2 Fucoidan).
Notorious binding protein method or immuno method should not be used, since not so specific and non-quantitative. Famous biotin-binding protein of Avidin also has turned out to be the lipoic acid/thioctic acid- and amino-acids- binding protein.
- Badges
- Science method
- Similar topics
- Biological Science
- Genetics
- Gene Expression
- Protein Expression