Additional file 2 of Lessons from single cell omics: admixed American ancestry and sex confer cardiometabolic disease risk in Mexicans

Le résumé fourni par la source

Additional file 2: Supplementary Figures. Figs. S1-32. Fig. S1: Subcutaneous adipose tissue single nucleus RNA-sequencing cohort from Mexico City has a high average proportion of estimated global admixed American (AMR) ancestry. Fig. S2: Multi-step quality control of SAT snRNA-seq data from 49 Mexican individuals produces a large SAT single cell reference. Fig. S3: The contexts and cardiometabolic disease (CMD) traits show significant correlations in the Mexican study cohorts. Fig. S4: The proportions in the Mexican SAT snRNA-seq cohort of the four main cell-types are comparable to those of previously published SAT snRNA-seq cohorts. Fig. S5: Principal component analysis of cell-type level pseudobulk gene expression per sample. Fig. S6: Comparisons of the cell-type proportions derived from the single cell level data by context and CMD traits detect relatively minor differences. Fig. S7: The proportions of main cell-types and cellular subtypes differ by binary traits (sex and type 2 diabetes (T2D) status) and correlate with continuous contexts and CMD traits. Fig. S8: The Mexican SAT snRNA-seq data shows statistically significant differences by sex, BMI, and age on the reduced dimension space. Fig. S9: Single nucleus RNA-sequencing data of SAT biopsies from 49 Mexican individuals shows differences in the UMAP space by the CMD traits of type 2 diabetes (T2D), serum triglycerides (TGs), serum total cholesterol (TC), and serum HDL-cholesterol (HDL-C). Fig. S10: Multi-cellular factor analysis (MOFAcell) [78] reveals that particularly adipocytes display strong cellular heterogeneity, with sex and BMI influencing their variability. Fig. S11: Ancestry and total serum triglycerides affect gene expression in adipocytes and macrophages, respectively, and the differentially expressed genes by BMI, sex, and triglycerides show functional enrichments. Fig. S12: Adipose stem and precursor cells (ASPCs) contain functionally distinct subtypes and cell-type level co-expression networks. Fig. S13: Subtypes and cell-type level co-expression networks within the vascular cell-types. Fig. S14: The subtypes and cell-type level co-expression networks within the lymphoid cell-types capture distinct subtype functions. Fig. S15: Subtypes and cell-type level co-expression networks within the myeloid cell-types. Fig. S16: Quality of the Mexican SAT snRNA-seq data compared to that of the SAT snRNA-seq of a previously published adipose single cell atlas [5, 66]. Fig. S17: Comparisons of the identified subtypes per cell-type with SAT subtype annotations from published atlases [5, 21, 22, 66, 72, 73]. Fig. S18: The SAT cell-type level co-expression networks are highly preserved consistently across six independent external datasets. Fig. S19: Comparisons of the adipocyte subtype proportions by context and CMD traits detect sex differences among two adipocyte subtypes. Fig. S20: The ASPC subtype proportions exhibit minor differences by CMD context and CMD trait. Fig. S21: Proportions of the lymphoid subtypes show minor differences by CMD context and CMD trait. Fig. S22: Comparisons of myeloid subtypes identify a myeloid subtype associated with BMI. Fig. S23: The vascular subtypes show minor differences in proportions by CMD context and CMD trait. Fig. S24: The proportions of cellular subtypes differ by binary traits (sex and type 2 diabetes (T2D) status) and correlate with continuous contexts and CMD traits. Fig. S25: The proportions of cellular subtypes are correlated with continuous contexts and CMD traits. Fig. S26: Principal component analysis of the MetMex cohort on the 1000 Genomes PC space. Fig. S27: Power estimation analysis indicates that we have adequate power (>80%) to detect cell-type level cis-eQTLs with realistic effect sizes. Fig. S28: Cell-type level cis-eQTL variants differ in their allele frequencies between the Mexicans and non-Finnish Europeans from gnomAD [129]. Fig. S29: The ancestry-stratified ASPC, endothelial cell, and macrophage cis-eQTL variants overlap and colocalize with GWAS variants for serum triglycerides in the large trans-ancestry GWAS [88]. Fig. S30: The ancestry-stratified cell-type level cis-eQTL variants overlap and colocalize with GWAS variants for serum HDL-cholesterol in the large trans-ancestry GWAS [88]. Fig. S31: The ancestry-stratified cell-type level cis-eQTL variants overlap and colocalize with GWAS variants for serum total cholesterol in the large trans-ancestry GWAS [88]. Fig. S32: Of the colocalized genes, 25 genes have not been reported in previous SAT bulk colocalization studies [101], including 12 genes regulated by Mexican-enriched and 7 genes regulated by European enriched colocalized risk variants.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.