Fig. 8: Clinical group stratification by combining somatic mutation, DNA methylation and microbiome data in a mixed integration approach. | Nature Communications

Fig. 8: Clinical group stratification by combining somatic mutation, DNA methylation and microbiome data in a mixed integration approach.

From: Hologenomic analysis of rectal mucus sampling for detection of adenomatous polyps and colorectal cancer

Fig. 8: Clinical group stratification by combining somatic mutation, DNA methylation and microbiome data in a mixed integration approach.

a Heatmap for the tri-omics case (somatic mutation, DNA methylation and microbiome) showing clustering across biomarkers and participants (control n = 30; CRC n = 38; early-stage polyps = 7; late-stage polys = 4). Feature selection was performed before integration, leading to the 30 biomarkers listed in the plot. b Tri-omics PCA analysis showing the data projected onto pairs of principal components (control n = 30; CRC n = 38; early-stage polyps = 7; late-stage polys = 4). Polyps cases are transformed and plotted for comparison. The percent of explained variance for each PC is indicated in parentheses. c As in b, but for the dual-omics case with only mutation and DNA methylation data included (control n = 36; CRC n = 44; early-stage polyps = 12; late-stage polys = 4). E-Polyps + , Early-stage polyps; L-Polyps + , Late-stage polyps.

Back to article page