Fig. 4: Similar gene expression levels in different patients may be attributed to different unbalanced processes.

a Transcriptome-barcodes of two AML patients from the TARGET dataset (bone marrow samples), representing the patient-specific combination of unbalanced processes. These patients were classified as being in \({c}_{3}\) critical point by ST analysis. b The fold changes of six AML biomarkers (see supplementary methods) S100A9, CD34, FUT4, HOMER3, MEIS1, and JAG1 were up-regulated in both patients relative to their median expression levels across 190 other patients in the TARGET data set. c Detail of a network diagram of unbalanced processes, comprising the barcodes of patients 98 and 110 (a) and in which the selected biomarkers whose levels are most influenced by those processes. Green denotes up-regulation, red denotes down-regulation, and gray denotes no change. The size of the biomarker indicates its relative weight in each process. Functional connections are derived from the STRING database.