Figure 1.
Flowchart for individuals’ selection from the DiOGenes cohort.
Participants entering subsequent phases of the study as well as dropouts are indicated in total. AT, adipose tissue; CID, clinical investigation day; FA, fatty acids; HGI, high glycemic index; HP, high protein content; LCD, low calorie diet; LGI, low glycemic index; LP, low protein content; WMD, weight maintenance diet.
Figure 2.
Workflow of the network analysis.
Intra- and inter-omics networks were first inferred separately before a global merging for each time step: at baseline, after a 8-week low calorie diet and after weight maintenance diet (3 groups). Node clustering was then performed and clusters were systematically analyzed with most important nodes extraction and across time comparison. Resulting products of each step are given in purple and significant tests to assess the validity of the approach are given in yellow.
Table 1.
Baseline anthropometric and clinical characteristics of the 135 obese women.
Table 2.
Anthropometric and clinical characteristics of women according to weight control groups during dietary intervention.
Figure 3.
Baseline adipose tissue networks of obese women.
A sparse Graphical Gaussian Model (GGM) was used to estimate partial correlations in each set of variables and regularized canonical correlation analysis (CCA) was used to assess links between paired sets of variable. Clustering was performed using a spin glass model and simulated annealing. This analysis displays the variables that are connected independently from other variables. Graphs were laid out using force-based algorithms in Gephi 0.8.2 software. Nodes’ colors and font size indicate betweenness centrality. The red nodes have the highest betweenness and the green nodes the lowest one. Edge thickness is proportional to the strength of the correlation (CCA) or of the partial correlation (GGM). Edge color indicates the correlation sign: red for positive correlations and blue for negative ones. BMI, body mass index; CRP, C reactive protein; DBP, diastolic blood pressure; HDL, high density lipoprotein; LDL, low density lipoprotein; SBP, systolic blood pressure; TG, plasma triglycerides; waist, waist circumference.
Figure 4.
Adipose tissue networks of obese women after a 8-week low calorie diet.
A sparse Graphical Gaussian Model (GGM) was used to estimate partial correlations in each set of variables and regularized canonical correlation analysis (CCA) was used to assess links between paired sets of variable. Clustering was performed using a spin glass model and simulated annealing. This analysis displays the variables that are connected independently from other variables. Graphs were laid out using force-based algorithms in Gephi 0.8.2 software. Nodes’ colors and font size indicate betweenness centrality. The red nodes have the highest betweenness and the green nodes the lowest one. Edge thickness is proportional to the strength of the correlation (CCA) or of the partial correlation (GGM). Edge color indicates the correlation sign: red for positive correlations and blue for negative ones. BMI, body mass index; CRP, C reactive protein; DBP, diastolic blood pressure; HDL, high density lipoprotein; LDL, low density lipoprotein; SBP, systolic blood pressure; TG, plasma triglycerides; waist, waist circumference.
Figure 5.
Adipose tissue networks during maintenance phase in women regaining weight.
A sparse Graphical Gaussian Model (GGM) was used to estimate partial correlations in each set of variables and regularized canonical correlation analysis (CCA) was used to assess links between paired sets of variable. Clustering was performed using a spin glass model and simulated annealing. This analysis displays the variables that are connected independently from other variables. Graphs were laid out using force-based algorithms in Gephi 0.8.2 software. Nodes’ colors and font size indicate betweenness centrality. The red nodes have the highest betweenness and the green nodes the lowest one. Edge thickness is proportional to the strength of the correlation (CCA) or of the partial correlation (GGM). Edge color indicates the correlation sign: red for positive correlations and blue for negative ones. BMI, body mass index; CRP, C reactive protein; DBP, diastolic blood pressure; HDL, high density lipoprotein; LDL, low density lipoprotein; SBP, systolic blood pressure; TG, plasma triglycerides; waist, waist circumference.
Figure 6.
Adipose tissue networks during maintenance phase in women with continued weight loss.
A sparse Graphical Gaussian Model (GGM) was used to estimate partial correlations in each set of variables and regularized canonical correlation analysis (CCA) was used to assess links between paired sets of variable. Clustering was performed using a spin glass model and simulated annealing. This analysis displays the variables that are connected independently from other variables. Graphs were laid out using force-based algorithms in Gephi 0.8.2 software. Nodes’ colors and font size indicate betweenness centrality. The red nodes have the highest betweenness and the green nodes the lowest one. Edge thickness is proportional to the strength of the correlation (CCA) or of the partial correlation (GGM). Edge color indicates the correlation sign: red for positive correlations and blue for negative ones. BMI, body mass index; CRP, C reactive protein; DBP, diastolic blood pressure; HDL, high density lipoprotein; LDL, low density lipoprotein; SBP, systolic blood pressure; TG, plasma triglycerides; waist, waist circumference.