Figure 1
Gene expression overview of 105 soft tissue tumors. (a) Principal component analysis. A total of 12 599 probe sets with expression values not less than 100 in at least three samples were used in this analysis. x and y axes represent the first and second principal components (PC1 and PC2), respectively. Each dot represents a sample colored by its histological type. (b) Two-dimensional hierarchical clustering analysis. A total of 2590 probe sets differentially expressed among histological types (P<1.0 × 10−5 by ANOVA) were used. Columns represent samples and rows represent probe sets. Red and green indicate high and low expression, respectively. The 2590 probe sets were roughly divided into six clusters (clusters 1–6). The six graphs on the right show averages of normalized expression values of those clusters for each histological type. Note that spindle cell and pleomorphic sarcomas, such as dedifferentiated liposarcoma, myxofibrosarcoma, leiomyosarcoma, MPNST, fibrosarcoma and MFH, form a loose cluster and share a similar expression profile compared with synovial sarcoma, myxoid/round cell liposarcoma, well-differentiated liposarcoma and lipoma. SS, synovial sarcoma; MLS, myxoid/round cell liposarcoma; WDLS, well-differentiated liposarcoma; DDLS, dedifferentiated liposarcoma; MFS, myxofibrosarcoma; LMS, leiomyosarcoma and FS, fibrosarcoma.
