Fig. 9

The sensitivity and specificity of diagnostics based on the codon employment shift. a. We used transcriptomics data from the following studies: a lung cancer study for the first two graphs [27]; a multiple sclerosis study for the third graph [35]; data for the multiple cancer studies for the fourth graph [27,28,29,30]. Please refer to the main text for more information on how the diagnostic approach was designed. The bars show the sensitivity and specificity (as defined in medical diagnosis where the sensitivity is the ability of a test to properly recognize patients with the disease and the specificity is the ability of the test to properly recognize those free from the disease). b. The sensitivity (black) and the specificity (red) of the lung cancer diagnosis under conditions of randomized mRNA abundances. The mRNA abundance values in patients and controls were randomly changed, up or down, by a fold factor indicated on the X-axis. The same diagnostic analysis as in Fig. 9a then followed. Changing randomly all measurements by up to 10-fold leaves the precision of the diagnostics within the clinically accepted limits (85%)