Fig. 2

Similarity Network Fusion identifies clinically relevant clusters. A Placenta-by-placenta similarity matrix after similarity network fusion (SNF). Unsupervised clustering combining all molecular data types identified four clusters in the cohort (n = 271), demarcated with a dotted line. B A bar plot presenting the distribution of clinical cases across the SNF clusters. Left—numbers. Right—percentage. The bar on the right represents the expected proportions based on the entire study cohort. The table presents the percentage of placentas of each clinical group within each SNF cluster. *Different from all other clusters, **different from clusters III and IV. Chi-square test with a post hoc pairwise Fisher’s exact test. Post hoc Fisher’s exact test p-values are adjusted for multiple comparisons by the Holm-Bonferroni procedure. C A pie chart of the distribution early-onset PE cases across clusters. D The proportions of early- (< 34) vs late-onset PE (≥ 34 weeks) across the SNF clusters. Control-PT, control preterm; PE, preeclampsia; FGR, fetal growth restriction; FGR + HDP, fetal growth restriction with hypertensive disorder of pregnancy; sPTD, spontaneous preterm delivery