ELBOW - Evaluating foLd change By the lOgit Way
Description
Elbow an improved fold change test that uses cluster
analysis and pattern recognition to set cut off limits that are
derived directly from intrareplicate variance without assuming
a normal distribution for as few as 2 biological replicates.
Elbow also provides the same consistency as fold testing in
cross platform analysis. Elbow has lower false positive and
false negative rates than standard fold testing when both are
evaluated using T testing and Statistical Analysis of
Microarray using 12 replicates (six replicates each for initial
and final conditions). Elbow provides a null value based on
initial condition replicates and gives error bounds for results
to allow better evaluation of significance.