Skip to main content
. 2016 Jun 15;7:11743. doi: 10.1038/ncomms11743

Table 2. Percentile ranking of Spearman correlation among 1,082 genes.

Gene mutation MSI-H Distant-Mets Gene mutation
Gene signature
      APC TP53 KRAS BRAF EMT RAS
APC 0.1 99.6 100 100 0.1 23 0.2
TP53 0.2 99.4 99.9 6 0.4 74 0.1
KRAS 0.3 100 100 18 0.2 2 100
SMAD4 4 95 84 85 93 52 13 89
TCF7L2 22 58 98 67 70 8 49 29
FAM123B 36 13 97 31 97 22 71 80
GNAS 49 69 60 67 86 71 28 98
ADAMTS18 51 51 40 56 69 50 38 56
LRP1B 66 81 45 91 38 80 55 58
PIK3CA 67 28 85 7 98 32 1.4 51
TAF1L 81 48 39 80 21 93 24 84
FBXW7 86 12 63 56 38 78 41 54
SYNE1 93 21 4 30 50 95 24 99
CSMD3 93 34 7 48 62 99 6 99.4
CBX4 95 7 3 43 2 98 56 50
ITGB4 99.9 17 0.8 11 46 99.6 43 99.7
BRAF 100 3 0.1 16 0.1 37 99.9

Only genes with Z5 non-silent mutations were analysed. APC, Adenomatous polyposis coli; EMT, epithelial–mesenchymal transition; Mets, metastasis; MSI, microsatellite instability.

(1) Closer examination of the literature and our own data suggest that only the BRAF(V600E) (comprising 53/60 patients with BRAF mutations) and APC-truncating mutations (stopgains or frameshift mutations, as these mutations clearly inactivate downstream β-catenin-binding sites) should be considered as clearly functional for these two genes; hereafter, we include only BRAF(V600E) and APC-truncating mutations in the further analysis and simply refer to these specific mutations as BRAF and APC mutations.

(2) MSI-H is a binary variable for the MSI-H status.

(3) EMT and RAS refer to published gene signature scores16,17.

(4) Percentiles ≤5 or ⩾95 are shown in bold.