Rank-Based miRNA Signatures for Early Cancer Detection
Figure 2
Maps for the two groups of subjects: (a) Caucasian subjects and (b) African American subjects. Each node in the graph represents a subject, whose transcriptional signature was derived from a profile of circulating miRNA. The length of an edge is approximately proportional to the inverse of the distance between two signatures as computed by the algorithm. The nodes spontaneously cluster in well-defined and easily identifiable control/disease groups, with only two misclassifications in the Caucasian subjects case. Node label legend: red = early breast cancer subject; green = control subject.
