Figure 1.
Systematic analysis reveals heterogeneity in transcriptomic studies of young and aged murine HSCs. (A) Selection of young and aged HSC transcriptomic studies. Representation of decision tree used to acquire and analyze the different publications used for the systematic analysis. (B) Overview of the age groups used in each of the selected studies. The age group of mice used in each study (y-axis) is represented in months (x-axis). Color codes refer to the transcriptome platform used. Squares next to each study indicate in which kind of analysis each study was included. (C) Heterogeneity across different transcriptomic studies. Number of differentially expressed (DE) genes (DEGs) reported by each study. Left panel shows the number of DEGs as reported by the authors in their respective manuscripts. Right panel shows the number of DEGs identified by us using a uniform reanalysis. Red bars indicate the number of genes that are upregulated upon aging, while blue bars indicate downregulated genes. In the reanalysis panel, “h-Chambers” indicates half Chambers, meaning that just half of the DEGs were taken into account. (D) Ranking of DEGs across different studies. Horizontal bars represent the number of DEGs (x-axis, the actual number of DEG genes are shown in each bar), ranked by the number of studies in which they were found (occurrence in studies [consistency], y-axis). Orange and green bars indicate meta-analysis and reanalysis, respectively. (E) The list of genes gets more consistent with the increase number of studies added. Simulation of different list assembly using combinations of subsets of 12 studies (gradient color from blue to red). As the number of studies increased, the coefficient of variation (y-axis) decreases. Dots represent the different values and mean consistency for individual genes (see supplemental Data for further details). (F) Predicted resolution to discover consistently reported genes. The model demonstrates that >90% of all HSC aging genes have been identified (y-axis) in the 2 analytic approaches. Purple and green dashed lines indicate meta-analysis and reanalysis, respectively. (G) Extent of overlap between all independent studies and the HSC AS (centered red circle, AS). The size of the circles represents the number of DEGs identified in each study, the number of overlapping DEGs with the AS is represented next to each study identification, and the distance from the center represents the degree of similarity of each study to the HSC AS. Studies that show a higher degree of similarity are closer to the AS.

Systematic analysis reveals heterogeneity in transcriptomic studies of young and aged murine HSCs. (A) Selection of young and aged HSC transcriptomic studies. Representation of decision tree used to acquire and analyze the different publications used for the systematic analysis. (B) Overview of the age groups used in each of the selected studies. The age group of mice used in each study (y-axis) is represented in months (x-axis). Color codes refer to the transcriptome platform used. Squares next to each study indicate in which kind of analysis each study was included. (C) Heterogeneity across different transcriptomic studies. Number of differentially expressed (DE) genes (DEGs) reported by each study. Left panel shows the number of DEGs as reported by the authors in their respective manuscripts. Right panel shows the number of DEGs identified by us using a uniform reanalysis. Red bars indicate the number of genes that are upregulated upon aging, while blue bars indicate downregulated genes. In the reanalysis panel, “h-Chambers” indicates half Chambers, meaning that just half of the DEGs were taken into account. (D) Ranking of DEGs across different studies. Horizontal bars represent the number of DEGs (x-axis, the actual number of DEG genes are shown in each bar), ranked by the number of studies in which they were found (occurrence in studies [consistency], y-axis). Orange and green bars indicate meta-analysis and reanalysis, respectively. (E) The list of genes gets more consistent with the increase number of studies added. Simulation of different list assembly using combinations of subsets of 12 studies (gradient color from blue to red). As the number of studies increased, the coefficient of variation (y-axis) decreases. Dots represent the different values and mean consistency for individual genes (see supplemental Data for further details). (F) Predicted resolution to discover consistently reported genes. The model demonstrates that >90% of all HSC aging genes have been identified (y-axis) in the 2 analytic approaches. Purple and green dashed lines indicate meta-analysis and reanalysis, respectively. (G) Extent of overlap between all independent studies and the HSC AS (centered red circle, AS). The size of the circles represents the number of DEGs identified in each study, the number of overlapping DEGs with the AS is represented next to each study identification, and the distance from the center represents the degree of similarity of each study to the HSC AS. Studies that show a higher degree of similarity are closer to the AS.

Close Modal

or Create an Account

Close Modal
Close Modal