Reliability Measures

Overview

A reliability measure in GeneLinker™ is a numerical indication of the quality or reliability of a the measurement of an individual gene’s expression in an individual sample. GeneLinker™ expects reliability measures to fall between 0 and 1, with 0 representing very reliable and 1 representing unreliable. This is patterned off the interpretation of p-values in traditional statistical tests, where small numbers indicate significance.

Reliability measures can come from several sources:

Some microarray analysis programs can generate an estimate of the measurement of each spot on each chip. For example, Affymetrix MAS 4.0 can export a Call with a value of Present (P), Marginal (M), or Absent (A) for each spot. Affymetrix MAS 5.0 can export a Detection p-value which lies between zero (definitely present) and one (definitely absent).

If you have microarray data which replicates genes on a single chip, some of GeneLinker™’s import templates can convert those replicated values into a merged (averaged) value and an associated reliability measure. See Merging Within-Chip Replicate Measurements for more information.

Finally, you can generate reliability measures yourself in tabular format and import them in concert with tabular data by choosing the Tabular With Reliability Measures import template.