Fig. 2

Identification of the cord blood metabolites with altered abundance associated with anti-TPO antibodies positivity. a Hierarchical clustering of the top 79 1H-NMR features with altered abundances in the anti-TPO antibodies positivity group. Euclidean distance and the Ward method were used to generate clustering tree. Abundances of the 1H-NMR features were displayed as colors ranging from red to blue as shown in the color bar (red, higher abundance, blue; lower abundance). b Variable Importance (VIP) plot of the top 39 1H-NMR features with VIP > 1. The horizontal axis represents the 1H-NMR features; the vertical axis represents the values of the corresponding 1H-NMR features. The columns were colored according to the values of spearman rank correlation coefficient as shown in the color bar. The detailed VIP values are listed in Table S1. c S-Plot constructed from the OPLS-DA model shown in Fig. 1b, showing the covariance p [1] against the correlation coefficient p(corr) [1] of the 1H-NMR features with altered abundances in the Anti-TPO antibodies positivity group. The solid cycles were colored according to the values of spearman rank correlation coefficient as shown in the color bar. The identified metabolites with altered abundance in the anti-TPO antibodies group were labeled. The corresponding data are listed in Table S2. d Volcano plot of the 1H-NMR features with between-group altered abundances. The horizontal axis displays the log twofold change (FC) value, and the vertical axis represents the value of log 10 FDR-adjusted p value. The thresholds were set at FC ≥ 2, and FDR-adjusted P < 0.05. The blue and red circles denote the down- and up-regulated 1H-NMR features in the anti-TPO antibodies positivity group, respectively. The grey circles denote the 1H-NMR features without marked differences. The corresponding data are listed in Table S3