Fig. 4From: Knowledge graph construction based on granulosa cells transcriptome from polycystic ovary syndrome with normoandrogen and hyperandrogenIdentifying HA PCOS and NA PCOS Classification Markers and Using Classification Markers to Integrate with downloaded PCOS transcriptome data. A Flowchart for the process of identifying HA PCOS and NA PCOS Classification Markers. B The selected classification markers for HA PCOS, NA PCOS and Normal (NM) groups. C The classification precision for downloaded datasets of GSE34526, GSE102293 and GSE98595 (blue) and our data GSE137684 (red). D The logarithm of mean co-occurrence counts of HA PCOS and NA PCOS up-regulated genes after merging with metabolic anomaly correlated with PCOS. E, F The average p value (E) and hazard ratio (F) for Kaplan–Meier Suivival Anaysis on the 4 female specific cancers (BRCA = breast cancer, OV = ovarian cancer, UCEC = endometrial cancer, UCS = uterine cancer) for HA PCOS and NA PCOS up-regulated genesBack to article page