Supplementary materials

<b>Supplementary Figure</b><b>：</b><b>Supplementary Fig. 1 </b>Heterogeneity of the 4 immune cell phenotypes associated with head and neck cancer. Funnel plot:(A) TD CD4+ _CD4+;(B) CD3 on CD39+ secreting Treg;(C) CD80 on granulocyte,(D) CD45 on CD33br HLA DR+ CD14dim.<b>Supplementary Fig. 2 </b>Leave-one-out sensitivity analysis of 4 immune cell phenotypes associated with head and neck cancer. (A) TD CD4+ _CD4+;(B) CD3 on CD39+ secreting Treg;(C) CD80 on granulocyte;(D) CD45 on CD33br HLA DR+ CD14dim.<b>Supplementary Fig. 3 </b>Heterogeneity of the 12 plasma metabolites associated with head and neck cancer. Funnel plot: (A) Alpha− ketoglutarate to pyruvate ratio;(B) Proline to trans−4− hydroxyproline ratio;(C) 3− hydroxypyridine glucuronide levels;(D) 1−(1−enyl−stearoyl)−2−linoleoyl−GPE (p−18:0/18:2);(E) Glucose to maltose ratio;(F) Glutamate levels;(G) 3−amino−2−pipe ridone levels;(H) Cortolone glucuronide (1) levels;(I) 4−allylphenol sulfate levels;(J) 7−alpha− hydroxy−3− oxo−4−cholestenoate (7−hoca) levels;(K) Stearidonate (18:4n3) levels;(L) 1,2−dipalmitoyl−gpc (16:0/16:0) levels.<b>Supplementary Fig. 4 </b>Leave-one-out sensitivity analysis of 12 plasma metabolites associated with head and neck cancer. (A) Alpha− ketoglutarate to pyruvate ratio;(B) Proline to trans−4− hydroxyproline ratio;(C) 3− hydroxypyridine glucuronide levels;(D) 1−(1−enyl−stearoyl)−2−linoleoyl−GPE (p−18:0/18:2) ;(E) Glucose to maltose ratio;(F) Glutamate levels;(G) 3−amino−2−pipe ridone levels;(H) Cortolone glucuronide (1) levels;(I) 4−allylphenol sulfate levels;(J) 7−alpha− hydroxy−3− oxo−4−cholestenoate (7−hoca) levels;(K) Stearidonate (18:4n3) levels;(L) 1,2−dipalmitoyl−gpc (16:0/16:0) levels.<b>Supplementary Table</b><b>：</b><b>Supplementary Table 1 </b>Heterogeneity of the 3 immune cell phenotypes associated with head and neck cancer.<b>Supplementary Table 2 </b>Pleiotropy of the 3 immune cell phenotypes associated with head and neck cancer.<b>Supplementary Table 3 </b>Heterogeneity of the 12 plasma metabolites associated with head and neck cancer.<b>Supplementary Table 4 </b>Pleiotropy of the 12 plasma metabolites associated with head and neck cancer.<b>Supplementary Table 5 </b>Mendelian randomization analyses of the causal effects between immune cells, plasma metabolites(β1).<b>Supplementary Table 6 </b>Mendelian randomization analyses of the causal effects between plasma metabolites and head and neck cancer(β2).<b>Supplementary Table 7 </b>Mendelian randomization analyses of the causal effects between immune cells, plasma metabolites and head and neck cancer.