Results of RHINO Analysis ========================= Results for Composite Single-Gene Predictors -------------------------------------------- Early Predictor: The best ternary logic predictor synthesized, which consists in average of 3.89 single gene-predictors, makes prediction in 16 out of 19 cases with 1 error (which directly translates into an error rate of 6%). The normalized weighted error rate (based on an equivalent binary predictor) is 13%. The p-value determined using 100 randomization tests is ~ 0.01. The best predictor based on all 19 subjects (i.e. without crossvalidation) contains exactly the following four 4 genes HBD, PI3, DHX58 (all positive), and IGKC (negative). Middle Predictor: The best predictor consists in average of 2.16 genes and makes predictions in 16 out of 19 cases with 3 errors (yielding a direct error rate of 19%). The normalized error rate is 23% and the p-value is ~ 0.06. The best predictor based on all 19 subjects consists of two genes: PI3 (positive) and HMBOX1 (negative). Late Predictor: The best predictor consists in average of 2.63 genes and makes predictions in 15 out of 19 cases with 1 error (translating into a direct error rate of 7%). The normalized error rate is 16% with a p-value of ~ 0.01. The best predictor based on all 19 subjects consisten of two genes: HMBOX1 (negative) and ASGR2 (positive). Relevant Log-File Fragments The following fragments from the log file contain the detailed results. Testing top composite predictors at time 6.0 ... Average Best Composite Predictor: Top-3.89 Validation results for top composite predictors ... CompPred Best - Sub: 19 Pred: 16 (0.84) Err: 1 (0.06) NErr: 0.13 WErr: 0.06 NWErr: 0.13 CScore: 0.13 Ok: 1,2,3,4,5,6,7,8,9,10,13,15,17,18,20 Err: 16 CompPred Top-1 - Sub: 19 Pred: 8 (0.42) Err: 2 (0.25) NErr: 0.39 WErr: 0.24 NWErr: 0.39 CScore: 0.39 Ok: 1,2,3,8,15,17 Err: 16,20 CompPred Top-2 - Sub: 19 Pred: 16 (0.84) Err: 3 (0.19) NErr: 0.24 WErr: 0.18 NWErr: 0.22 CScore: 0.22 Ok: 1,2,3,4,5,6,8,10,12,13,15,17,18 Err: 7,9,16 CompPred Top-3 - Sub: 19 Pred: 13 (0.68) Err: 1 (0.08) NErr: 0.21 WErr: 0.07 NWErr: 0.20 CScore: 0.20 Ok: 1,2,3,4,5,6,8,10,13,15,17,18 Err: 16 CompPred Top-4 - Sub: 17 Pred: 16 (0.94) Err: 1 (0.06) NErr: 0.09 WErr: 0.06 NWErr: 0.13 CScore: 0.13 Ok: 1,2,3,4,5,6,7,8,9,10,13,15,17,18,20 Err: 16 Randomization Summary ... Nullhypothesis: computed predictor can predict random classification with error rate <= x Composite Predictor Best - Tests: 1000 - p(e <= 0.05) = 0.00 p(e <= 0.10) = 0.00 p(e <= 0.15) = 0.00 p(e <= 0.20) = 0.02 p(e <= 0.25) = 0.05 p(e <= 0.30) = 0.12 p(e <= 0.35) = 0.20 p(e <= 0.40) = 0.29 p(e <= 0.45) = 0.42 p(e <= 0.50) = 0.55 Computing top composite predictors at time 6.0 (without crossvalidation) ... Computing predictors at time 6.0 ... Defining composite predictors ... 1. Pred HBD - Pat: Pos - Sub: 19 Pred: 10 (0.53) Err: 1 (0.10) NErr: 0.29 WErr: 0.09 NWErr: 0.29 Ok: 1,2,3,6,8,13,15,17,20 Err: 16 2. Pred PI3 - Pat: Pos - Sub: 19 Pred: 15 (0.79) Err: 4 (0.27) NErr: 0.32 WErr: 0.25 NWErr: 0.30 Ok: 1,2,3,4,5,8,10,12,15,17,18 Err: 7,9,16,20 3. Pred DHX58 - Pat: Pos - Sub: 19 Pred: 10 (0.53) Err: 3 (0.30) NErr: 0.39 WErr: 0.29 NWErr: 0.39 Ok: 2,5,7,8,9,11,18 Err: 3,15,16 4. Pred IGKC - Pat: Neg - Sub: 19 Pred: 6 (0.32) Err: 1 (0.17) NErr: 0.39 WErr: 0.18 NWErr: 0.40 Ok: 6,7,9,19,20 Err: 12 Testing top composite predictors at time 6.0 ... Best Composite Predictor: Top-4 Score: 0.08 Testing top composite predictors at time 21.0 ... Average Best Composite Predictor: Top-2.16 Validation results for top composite predictors ... CompPred Best - Sub: 19 Pred: 16 (0.84) Err: 3 (0.19) NErr: 0.24 WErr: 0.18 NWErr: 0.23 CScore: 0.23 Ok: 1,2,3,5,6,8,10,13,15,16,17,18,19 Err: 9,11,20 CompPred Top-1 - Sub: 19 Pred: 16 (0.84) Err: 4 (0.25) NErr: 0.29 WErr: 0.24 NWErr: 0.27 CScore: 0.27 Ok: 1,2,3,5,8,10,12,13,15,16,17,18 Err: 4,7,11,20 CompPred Top-2 - Sub: 19 Pred: 15 (0.79) Err: 3 (0.20) NErr: 0.26 WErr: 0.19 NWErr: 0.25 CScore: 0.25 Ok: 1,2,5,6,8,10,13,15,16,17,18,19 Err: 9,11,20 CompPred Top-3 - Sub: 4 Pred: 4 (1.00) Err: 2 (0.50) NErr: 0.50 WErr: 0.49 NWErr: 0.50 CScore: 0.50 Ok: 3,8 Err: 9,20 Randomization Summary ... Nullhypothesis: computed predictor can predict random classification with error rate <= x Composite Predictor Best - Tests: 1000 - p(e <= 0.05) = 0.00 p(e <= 0.10) = 0.00 p(e <= 0.15) = 0.00 p(e <= 0.20) = 0.02 p(e <= 0.25) = 0.06 p(e <= 0.30) = 0.12 p(e <= 0.35) = 0.20 p(e <= 0.40) = 0.30 p(e <= 0.45) = 0.41 p(e <= 0.50) = 0.54 Computing top composite predictors at time 21.0 (without crossvalidation) ... Computing predictors at time 21.0 ... Defining composite predictors ... 1. Pred PI3 - Pat: Pos - Sub: 19 Pred: 16 (0.84) Err: 3 (0.19) NErr: 0.24 WErr: 0.18 NWErr: 0.22 Ok: 1,2,3,4,5,8,10,12,13,15,16,17,18 Err: 7,11,20 2. Pred HMBOX1 - Pat: Neg - Sub: 19 Pred: 13 (0.68) Err: 2 (0.15) NErr: 0.26 WErr: 0.16 NWErr: 0.27 Ok: 2,3,6,7,8,9,13,15,16,19,20 Err: 4,12 Testing top composite predictors at time 21.0 ... Best Composite Predictor: Top-2 Score: 0.15 Testing top composite predictors at time 66.0 ... Average Best Composite Predictor: Top-2.63 Validation results for top composite predictors ... CompPred Best - Sub: 19 Pred: 15 (0.79) Err: 1 (0.07) NErr: 0.16 WErr: 0.07 NWErr: 0.16 CScore: 0.16 Ok: 1,2,3,4,6,7,9,10,11,13,15,16,18,20 Err: 12 CompPred Top-1 - Sub: 19 Pred: 9 (0.47) Err: 2 (0.22) NErr: 0.37 WErr: 0.23 NWErr: 0.38 CScore: 0.38 Ok: 2,3,6,7,9,15,16 Err: 13,19 CompPred Top-2 - Sub: 19 Pred: 10 (0.53) Err: 1 (0.10) NErr: 0.29 WErr: 0.11 NWErr: 0.30 CScore: 0.30 Ok: 2,3,6,7,9,10,11,15,16 Err: 12 CompPred Top-3 - Sub: 19 Pred: 17 (0.89) Err: 2 (0.12) NErr: 0.16 WErr: 0.12 NWErr: 0.17 CScore: 0.17 Ok: 1,2,3,4,6,7,9,10,11,13,15,16,18,19,20 Err: 5,12 CompPred Top-4 - Sub: 13 Pred: 13 (1.00) Err: 1 (0.08) NErr: 0.08 WErr: 0.08 NWErr: 0.22 CScore: 0.22 Ok: 2,3,6,7,8,9,10,11,15,16,19,20 Err: 12 Computing top composite predictors at time 6.0 (without crossvalidation) ... Randomization Summary ... Nullhypothesis: computed predictor can predict random classification with error rate <= x Composite Predictor Best - Tests: 1000 - p(e <= 0.05) = 0.00 p(e <= 0.10) = 0.00 p(e <= 0.15) = 0.00 p(e <= 0.20) = 0.01 p(e <= 0.25) = 0.04 p(e <= 0.30) = 0.07 p(e <= 0.35) = 0.15 p(e <= 0.40) = 0.26 p(e <= 0.45) = 0.40 p(e <= 0.50) = 0.53 Computing top composite predictors at time 66.0 (without crossvalidation) ... Computing predictors at time 66.0 ... Defining composite predictors ... 1. Pred HMBOX1 - Pat: Neg - Sub: 19 Pred: 11 (0.58) Err: 0 (0.00) NErr: 0.21 WErr: 0.00 NWErr: 0.22 Ok: 2,3,6,7,8,9,13,15,16,19,20 Err: 2. Pred ASGR2 - Pat: Pos - Sub: 19 Pred: 13 (0.68) Err: 2 (0.15) NErr: 0.26 WErr: 0.15 NWErr: 0.26 Ok: 1,2,3,4,7,10,11,13,16,18,20 Err: 12,19 3. Pred CXCL10 - Pat: Pos - Sub: 19 Pred: 9 (0.47) Err: 2 (0.22) NErr: 0.37 WErr: 0.24 NWErr: 0.38 Ok: 3,6,7,11,15,16,19 Err: 5,13 4. Pred PI3 - Pat: Pos - Sub: 19 Pred: 12 (0.63) Err: 5 (0.42) NErr: 0.45 WErr: 0.42 NWErr: 0.45 Ok: 3,4,6,15,16,18,19 Err: 7,11,13,17,20 Testing top composite predictors at time 66.0 ... Best Composite Predictor: Top-2 Score: 0.14 Results for Composite Single-Gene Predictors with Preselected Genes ------------------------------------------------------------------- Early Predictor: The best ternary logic predictor synthesized, which consists in average of 3.16 single gene-predictors, makes prediction in 13 out of 19 cases with 5 errors (which directly translates into an error rate of 38%). The normalized weighted error rate (based on an equivalent binary predictor) is 41%. The p-value determined using 100 randomization tests is ~ 0.26. The best predictor based on all 19 subjects (i.e. without crossvalidation) contains exactly the following three genes: HLA-DRB1 (negative), DHX58 (positive), and CTSG (negative). Middle Predictor: The best predictor consists in average of 1.00 genes and makes predictions in 13 out of 19 cases with 1 errors (yielding a direct error rate of 8%). The normalized error rate is 22% and the p-value is ~ 0.13. The best predictor based on all 19 subjects consists of one gene: NUP85 (negative). Late Predictor: The best predictor consists in average of 1.89 genes and makes predictions in 13 out of 19 cases with 4 error (translating into a direct error rate of 31%). The normalized error rate is 38% with a p-value of ~ 0.27. The best predictor based on all 19 subjects consisten of three genes: IFITM1 (positive), PRKRA, and CEACAM8 (both negative). Relevant Log-File Fragments Testing top composite predictors at time 6.0 ... Average Best Composite Predictor: Top-3.16 Validation results for top composite predictors ... CompPred Best - Sub: 19 Pred: 13 (0.68) Err: 5 (0.38) NErr: 0.42 WErr: 0.37 NWErr: 0.41 CScore: 0.41 Ok: 2,5,6,8,11,13,18,20 Err: 3,7,10,15,16 CompPred Top-1 - Sub: 19 Pred: 8 (0.42) Err: 7 (0.88) NErr: 0.66 WErr: 0.88 NWErr: 0.66 CScore: 0.66 Ok: 7 Err: 3,9,10,11,15,16,18 CompPred Top-2 - Sub: 19 Pred: 9 (0.47) Err: 6 (0.67) NErr: 0.58 WErr: 0.66 NWErr: 0.58 CScore: 0.58 Ok: 2,6,20 Err: 3,9,10,11,15,16 CompPred Top-3 - Sub: 19 Pred: 13 (0.68) Err: 6 (0.46) NErr: 0.47 WErr: 0.45 NWErr: 0.46 CScore: 0.46 Ok: 2,5,6,8,13,18,20 Err: 3,7,9,10,15,16 CompPred Top-4 - Sub: 6 Pred: 4 (0.67) Err: 2 (0.50) NErr: 0.50 WErr: 0.49 NWErr: 0.50 CScore: 0.50 Ok: 11,18 Err: 3,16 CompPred Top-5 - Sub: 2 Pred: 2 (1.00) Err: 1 (0.50) NErr: 0.50 WErr: 0.50 NWErr: 0.50 CScore: 0.50 Ok: 11 Err: 3 Randomization Summary ... Nullhypothesis: computed predictor can predict random classification with error rate <= x Composite Predictor Best - Tests: 100 - p(e <= 0.05) = 0.00 p(e <= 0.10) = 0.00 p(e <= 0.15) = 0.00 p(e <= 0.20) = 0.01 p(e <= 0.25) = 0.02 p(e <= 0.30) = 0.07 p(e <= 0.35) = 0.13 p(e <= 0.40) = 0.23 p(e <= 0.45) = 0.34 p(e <= 0.50) = 0.49 Computing top composite predictors at time 6.0 (without crossvalidation) ... Defining composite predictors ... 1. Pred HLA-DRB1 - Pat: Neg - Sub: 19 Pred: 8 (0.42) Err: 1 (0.12) NErr: 0.34 WErr: 0.12 NWErr: 0.35 Ok: 3,4,6,7,13,16,20 Err: 9 2. Pred DHX58 - Pat: Pos - Sub: 19 Pred: 11 (0.58) Err: 3 (0.27) NErr: 0.37 WErr: 0.26 NWErr: 0.36 Ok: 2,5,7,8,9,10,11,18 Err: 3,15,16 3. Pred CTSG - Pat: Neg - Sub: 19 Pred: 9 (0.47) Err: 2 (0.22) NErr: 0.37 WErr: 0.21 NWErr: 0.36 Ok: 1,2,7,9,10,16,18 Err: 3,15 Testing top composite predictors at time 6.0 ... Best Composite Predictor: Top-3 Score: 0.18 Testing top composite predictors at time 21.0 ... Average Best Composite Predictor: Top-1.00 Validation results for top composite predictors ... CompPred Best - Sub: 13 Pred: 13 (1.00) Err: 1 (0.08) NErr: 0.08 WErr: 0.08 NWErr: 0.22 CScore: 0.22 Ok: 2,3,4,6,7,8,9,11,15,16,19,20 Err: 13 CompPred Top-1 - Sub: 13 Pred: 13 (1.00) Err: 1 (0.08) NErr: 0.08 WErr: 0.08 NWErr: 0.22 CScore: 0.22 Ok: 2,3,4,6,7,8,9,11,15,16,19,20 Err: 13 Computing top composite predictors at time 66.0 (with crossvalidation) ... Randomization Summary ... Nullhypothesis: computed predictor can predict random classification with error rate <= x Composite Predictor Best - Tests: 100 - p(e <= 0.05) = 0.00 p(e <= 0.10) = 0.00 p(e <= 0.15) = 0.04 p(e <= 0.20) = 0.09 p(e <= 0.25) = 0.13 p(e <= 0.30) = 0.18 p(e <= 0.35) = 0.28 p(e <= 0.40) = 0.35 p(e <= 0.45) = 0.41 p(e <= 0.50) = 0.62 Computing top composite predictors at time 21.0 (without crossvalidation) ... Defining composite predictors ... 1. Pred NUP85 - Pat: Neg - Sub: 13 Pred: 13 (1.00) Err: 1 (0.08) NErr: 0.08 WErr: 0.08 NWErr: 0.22 Ok: 2,3,4,6,7,8,9,11,15,16,19,20 Err: 13 Testing top composite predictors at time 21.0 ... Best Composite Predictor: Top-1 Score: 0.22 Testing top composite predictors at time 66.0 ... Average Best Composite Predictor: Top-1.89 Validation results for top composite predictors ... CompPred Best - Sub: 19 Pred: 13 (0.68) Err: 4 (0.31) NErr: 0.37 WErr: 0.33 NWErr: 0.38 CScore: 0.38 Ok: 3,6,7,8,11,15,16,19,20 Err: 5,12,13,17 CompPred Top-1 - Sub: 19 Pred: 12 (0.63) Err: 3 (0.25) NErr: 0.34 WErr: 0.27 NWErr: 0.36 CScore: 0.36 Ok: 3,6,7,8,11,15,16,19,20 Err: 5,12,13 CompPred Top-2 - Sub: 19 Pred: 16 (0.84) Err: 6 (0.38) NErr: 0.39 WErr: 0.40 NWErr: 0.41 CScore: 0.41 Ok: 3,4,6,7,8,11,15,16,19,20 Err: 1,5,12,13,17,18 CompPred Top-3 - Sub: 9 Pred: 6 (0.67) Err: 3 (0.50) NErr: 0.50 WErr: 0.52 NWErr: 0.51 CScore: 0.51 Ok: 6,8,19 Err: 5,13,17 CompPred Top-4 - Sub: 1 Pred: 1 (1.00) Err: 1 (1.00) NErr: 1.00 WErr: 1.00 NWErr: 0.53 CScore: 0.53 Ok: Err: 13 Randomization Summary ... Nullhypothesis: computed predictor can predict random classification with error rate <= x Composite Predictor Best - Tests: 100 - p(e <= 0.05) = 0.00 p(e <= 0.10) = 0.00 p(e <= 0.15) = 0.00 p(e <= 0.20) = 0.01 p(e <= 0.25) = 0.07 p(e <= 0.30) = 0.14 p(e <= 0.35) = 0.19 p(e <= 0.40) = 0.27 p(e <= 0.45) = 0.38 p(e <= 0.50) = 0.55 Computing top composite predictors at time 66.0 (without crossvalidation) ... Defining composite predictors ... 1. Pred IFITM1 - Pat: Pos - Sub: 19 Pred: 10 (0.53) Err: 0 (0.00) NErr: 0.24 WErr: 0.00 NWErr: 0.25 Ok: 3,4,6,7,8,11,15,16,19,20 Err: 2. Pred PRKRA - Pat: Neg - Sub: 19 Pred: 9 (0.47) Err: 2 (0.22) NErr: 0.37 WErr: 0.24 NWErr: 0.38 Ok: 3,4,7,11,15,16,20 Err: 13,17 3. Pred CEACAM8 - Pat: Neg - Sub: 19 Pred: 4 (0.21) Err: 1 (0.25) NErr: 0.45 WErr: 0.24 NWErr: 0.44 Ok: 1,18,20 Err: 7 Testing top composite predictors at time 66.0 ... Best Composite Predictor: Top-3 Score: 0.25 Preliminary Results for Composite Cluster Predictors ---------------------------------------------------- Early Predictors: The best computed ternary predictor consists in average of 4.95 (sub)clusters. It makes predictions in 17 out of 19 cases with 6 errors (translating into 35% error rate). The normalized error rate is 37% and the p-value, determined using 100 randomization tests, is ~ 0.29. The best predictor based on all 19 subjects consists of 5 (sub)clusters named 0.16 (positive), 8.0 (negative), 0.46 (positive),7.0 (negative), and 0.22 (positive). See table below for the corresponding sets of genes. Middle Predictors: The best ternary predictor consists in average of 3.53 (sub)clusters. It makes predictions in 16 out of 19 cases with 5 errors (yielding an error rate of 31%). The normalized error rate is 34% and the p-value is ~ 0.20. The best predictor based on all 19 subjcts has 4 (sub)clusters: 11.0, 0.14, 7.0 (all positive), 0.44 (negative). Late Predictors: The best ternary predictor consists in average of 4.32 (sub)clusters. It makes predictions in 14 out of 19 cases with 3 errors (yielding an error rate of 21%). The normalized error rate is 30% with a p-value of ~ 0.05. The best predictor based on all 19 subjects has four (sub)clusters: 0.4 (positive), 11.1 (negative), 0.30 (negative), 8.0 (positive). Table of Relevant Clusters subcluster 0.4: { CNPY2 CETN2 PARK7 CSTA SEC11A METTL5 CHMP4A DNAJC15 H2AFZ ANXA2 ILF2 NDUFB8 NEDD8 CLEC4A POMP MRPS17 UFC1 PDHB ATP5J PPIB PSMD4 RPL36AL RPS6 SNRPD2 TTC1 TMEM14B DPM1 GADD45GIP1 RBM39 } subcluster 0.14: { CEBPG DBF4 GLT8D1 C2orf47 MINPP1 } subcluster 0.16: { CNPY2 EIF3M CEBPG CETN2 PTGES3 DBF4 DSTN PARK7 CSTA SEC11A HIBCH C6orf66 METTL5 CHMP4A DNAJC15 SLC25A5 H2AFZ ANXA2 HMGN1 ILF2 NDUFB8 NEDD8 CLEC4A POP5 POMP MRPS17 UFC1 PDHB ATP5J PPIB GLT8D1 PSMD4 PTGER2 RFC4 RPL36AL RPS6 MRPL9 SNRPD2 TPD52 TTC1 C2orf47 TMEM14B DPM1 GADD45GIP1 MINPP1 RBM39 } subcluster 0.22: { IGLV3-25 IGLV3-19 IGLV3-10 IGLV2-14 IGLV1-40 IGKV4-1 IGKV3-20 IGKC TXNDC5 IGLL3 } subcluster 0.30: { DIDO1 NISCH VAMP1 ZNF767 } subcluster 0.44: { PDZK1IP1 MYL9 KLF1 GPR44 RNF187 MKRN1 FKBP8 C18orf10 PLEK2 GP9 GUK1 GYPB HAGH HBD HBQ1 KRT1 LTBP1 GLRX5 ERAF PDK2 PPP2R5B BPGM CA1 UBXD1 SHARPIN SELENBP1 CDC34 } subcluster 0.46: { PDZK1IP1 MYL9 KLF1 GPR44 RNF187 F13A1 MKRN1 FKBP8 C18orf10 PLEK2 GP9 GUK1 GYPB HAGH HBD HBQ1 KRT1 LTBP1 GLRX5 ERAF PDK2 PF4 PPP2R5B BPGM CA1 UBXD1 SHARPIN SELENBP1 CDC34 } cluster 7.0: { TMEM176B TMEM176A } cluster 8.0: { IFI30 NAGK } cluster 11.0: { ZBTB16 DLG5 } subcluster 11.1: { CENTG2 ZBTB16 DLG5 } Relevant Log-File Fragments Testing top composite predictors at time 6.0 ... Average Best Composite Predictor: Top-4.95 Validation results for top composite predictors ... CompPred Best - Sub: 19 Pred: 17 (0.89) Err: 6 (0.35) NErr: 0.37 WErr: 0.35 NWErr: 0.37 CScore: 0.37 Ok: 2,3,4,5,6,12,13,15,17,19,20 Err: 7,9,10,11,16,18 CompPred Top-1 - Sub: 19 Pred: 4 (0.21) Err: 2 (0.50) NErr: 0.50 WErr: 0.49 NWErr: 0.50 CScore: 0.50 Ok: 2,17 Err: 16,18 CompPred Top-2 - Sub: 19 Pred: 5 (0.26) Err: 3 (0.60) NErr: 0.53 WErr: 0.59 NWErr: 0.52 CScore: 0.52 Ok: 2,17 Err: 10,16,18 CompPred Top-3 - Sub: 19 Pred: 12 (0.63) Err: 5 (0.42) NErr: 0.45 WErr: 0.41 NWErr: 0.44 CScore: 0.44 Ok: 2,3,4,5,13,17,19 Err: 6,10,16,18,20 CompPred Top-4 - Sub: 19 Pred: 14 (0.74) Err: 4 (0.29) NErr: 0.34 WErr: 0.28 NWErr: 0.34 CScore: 0.34 Ok: 1,2,3,4,5,12,13,15,17,19 Err: 7,10,16,18 CompPred Top-5 - Sub: 19 Pred: 19 (1.00) Err: 6 (0.32) NErr: 0.32 WErr: 0.31 NWErr: 0.31 CScore: 0.31 Ok: 1,2,3,4,5,6,8,12,13,15,17,19,20 Err: 7,9,10,11,16,18 CompPred Top-6 - Sub: 15 Pred: 14 (0.93) Err: 5 (0.36) NErr: 0.37 WErr: 0.35 NWErr: 0.39 CScore: 0.39 Ok: 1,2,3,4,5,12,13,15,20 Err: 9,10,11,16,18 Randomization Summary ... Nullhypothesis: computed predictor can predict random classification with error rate <= x Composite Predictor Best - Tests: 100 - p(e <= 0.05) = 0.00 p(e <= 0.10) = 0.00 p(e <= 0.15) = 0.01 p(e <= 0.20) = 0.01 p(e <= 0.25) = 0.05 p(e <= 0.30) = 0.12 p(e <= 0.35) = 0.19 p(e <= 0.40) = 0.29 p(e <= 0.45) = 0.39 p(e <= 0.50) = 0.60 Computing top composite predictors at time 6.0 (without crossvalidation) ... Computing predictors at time 6.0 ... Defining composite predictors ... 1. Pred 0.16 - Pat: Pos - Sub: 19 Pred: 7 (0.37) Err: 0 (0.00) NErr: 0.32 WErr: 0.00 NWErr: 0.31 Ok: 2,6,8,15,16,17,18 Err: 2. Pred 8.0 - Pat: Neg - Sub: 19 Pred: 10 (0.53) Err: 2 (0.20) NErr: 0.34 WErr: 0.20 NWErr: 0.34 Ok: 1,2,5,12,13,17,19,20 Err: 16,18 3. Pred 0.46 - Pat: Pos - Sub: 19 Pred: 8 (0.42) Err: 1 (0.12) NErr: 0.34 WErr: 0.12 NWErr: 0.34 Ok: 3,4,6,13,17,19,20 Err: 16 4. Pred 7.0 - Pat: Neg - Sub: 19 Pred: 8 (0.42) Err: 1 (0.12) NErr: 0.34 WErr: 0.13 NWErr: 0.35 Ok: 3,4,5,6,7,13,16 Err: 10 5. Pred 0.22 - Pat: Pos - Sub: 19 Pred: 12 (0.63) Err: 4 (0.33) NErr: 0.39 WErr: 0.31 NWErr: 0.38 Ok: 5,8,10,12,13,15,17,18 Err: 6,7,9,20 6. Pred 0.80 - Pat: Neg - Sub: 19 Pred: 10 (0.53) Err: 3 (0.30) NErr: 0.39 WErr: 0.29 NWErr: 0.39 Ok: 2,5,6,8,12,15,20 Err: 11,16,18 Testing top composite predictors at time 6.0 ... Best Composite Predictor: Top-5 Score: 0.15 0.16[0.15[0.12[0.4[0.0]],0.14]] { CNPY2 EIF3M CEBPG CETN2 PTGES3 DBF4 DSTN PARK7 CSTA SEC11A HIBCH C6orf66 METTL5 CHMP4A DNAJC15 SLC25A5 H2AFZ ANXA2 HMGN1 ILF2 NDUFB8 NEDD8 CLEC4A POP5 POMP MRPS17 UFC1 PDHB ATP5J PPIB GLT8D1 PSMD4 PTGER2 RFC4 RPL36AL RPS6 MRPL9 SNRPD2 TPD52 TTC1 C2orf47 TMEM14B DPM1 GADD45GIP1 MINPP1 RBM39 } 8.0 { IFI30 NAGK } 0.46[0.44[0.40[0.34[0.33[0.9[0.2]]]]]] { PDZK1IP1 MYL9 KLF1 GPR44 RNF187 F13A1 MKRN1 FKBP8 C18orf10 PLEK2 GP9 GUK1 GYPB HAGH HBD HBQ1 KRT1 LTBP1 GLRX5 ERAF PDK2 PF4 PPP2R5B BPGM CA1 UBXD1 SHARPIN SELENBP1 CDC34 } 7.0 { TMEM176B TMEM176A } 0.22[0.10] { IGLV3-25 IGLV3-19 IGLV3-10 IGLV2-14 IGLV1-40 IGKV4-1 IGKV3-20 IGKC TXNDC5 IGLL3 } Testing top composite predictors at time 21.0 ... Average Best Composite Predictor: Top-3.53 Validation results for top composite predictors ... CompPred Best - Sub: 19 Pred: 16 (0.84) Err: 5 (0.31) NErr: 0.34 WErr: 0.32 NWErr: 0.34 CScore: 0.34 Ok: 2,3,4,5,7,12,15,17,18,19,20 Err: 1,9,10,11,13 CompPred Top-1 - Sub: 19 Pred: 5 (0.26) Err: 5 (1.00) NErr: 0.63 WErr: 1.00 NWErr: 0.62 CScore: 0.62 Ok: Err: 4,9,11,15,16 CompPred Top-2 - Sub: 19 Pred: 11 (0.58) Err: 6 (0.55) NErr: 0.53 WErr: 0.53 NWErr: 0.52 CScore: 0.52 Ok: 2,3,5,12,18 Err: 1,4,9,10,11,16 CompPred Top-3 - Sub: 19 Pred: 13 (0.68) Err: 5 (0.38) NErr: 0.42 WErr: 0.38 NWErr: 0.42 CScore: 0.42 Ok: 2,3,5,12,15,17,18,19 Err: 1,9,10,11,13 CompPred Top-4 - Sub: 18 Pred: 16 (0.89) Err: 4 (0.25) NErr: 0.28 WErr: 0.25 NWErr: 0.29 CScore: 0.29 Ok: 2,3,4,5,6,7,12,15,17,18,19,20 Err: 1,9,11,13 CompPred Top-5 - Sub: 2 Pred: 2 (1.00) Err: 1 (0.50) NErr: 0.50 WErr: 0.50 NWErr: 0.50 CScore: 0.50 Ok: 4 Err: 11 Randomization Summary ... Nullhypothesis: computed predictor can predict random classification with error rate <= x Composite Predictor Best - Tests: 100 - p(e <= 0.05) = 0.00 p(e <= 0.10) = 0.00 p(e <= 0.15) = 0.00 p(e <= 0.20) = 0.01 p(e <= 0.25) = 0.08 p(e <= 0.30) = 0.14 p(e <= 0.35) = 0.20 p(e <= 0.40) = 0.32 p(e <= 0.45) = 0.45 p(e <= 0.50) = 0.63 Computing top composite predictors at time 21.0 (without crossvalidation) ... Computing predictors at time 21.0 ... Defining composite predictors ... 1. Pred 11.0 - Pat: Neg - Sub: 19 Pred: 12 (0.63) Err: 2 (0.17) NErr: 0.29 WErr: 0.16 NWErr: 0.29 Ok: 1,4,5,7,12,15,16,17,19,20 Err: 9,11 2. Pred 0.14 - Pat: Neg - Sub: 19 Pred: 8 (0.42) Err: 0 (0.00) NErr: 0.29 WErr: 0.00 NWErr: 0.29 Ok: 2,3,4,9,10,12,15,18 Err: 3. Pred 7.0 - Pat: Neg - Sub: 19 Pred: 11 (0.58) Err: 2 (0.18) NErr: 0.32 WErr: 0.17 NWErr: 0.31 Ok: 2,3,5,7,9,13,16,18,19 Err: 11,15 4. Pred 0.44 - Pat: Pos - Sub: 19 Pred: 11 (0.58) Err: 2 (0.18) NErr: 0.32 WErr: 0.18 NWErr: 0.32 Ok: 2,3,4,6,8,9,11,15,17 Err: 10,16 Testing top composite predictors at time 21.0 ... Best Composite Predictor: Top-4 Score: 0.08 11.0 { ZBTB16 DLG5 } 0.14 { CEBPG DBF4 GLT8D1 C2orf47 MINPP1 } 7.0 { TMEM176B TMEM176A } 0.44[0.40[0.34[0.33[0.9[0.2]]]]] { PDZK1IP1 MYL9 KLF1 GPR44 RNF187 MKRN1 FKBP8 C18orf10 PLEK2 GP9 GUK1 GYPB HAGH HBD HBQ1 KRT1 LTBP1 GLRX5 ERAF PDK2 PPP2R5B BPGM CA1 UBXD1 SHARPIN SELENBP1 CDC34 } Testing top composite predictors at time 66.0 ... Average Best Composite Predictor: Top-4.32 Validation results for top composite predictors ... CompPred Best - Sub: 19 Pred: 14 (0.74) Err: 3 (0.21) NErr: 0.29 WErr: 0.22 NWErr: 0.30 CScore: 0.30 Ok: 1,2,3,4,6,7,9,11,15,16,19 Err: 5,13,20 CompPred Top-1 - Sub: 19 Pred: 6 (0.32) Err: 2 (0.33) NErr: 0.45 WErr: 0.35 NWErr: 0.45 CScore: 0.45 Ok: 3,6,15,19 Err: 13,20 CompPred Top-2 - Sub: 19 Pred: 7 (0.37) Err: 2 (0.29) NErr: 0.42 WErr: 0.30 NWErr: 0.43 CScore: 0.43 Ok: 3,6,15,16,19 Err: 13,20 CompPred Top-3 - Sub: 19 Pred: 12 (0.63) Err: 3 (0.25) NErr: 0.34 WErr: 0.26 NWErr: 0.35 CScore: 0.35 Ok: 2,3,4,6,7,11,15,16,19 Err: 5,13,20 CompPred Top-4 - Sub: 18 Pred: 14 (0.78) Err: 3 (0.21) NErr: 0.28 WErr: 0.22 NWErr: 0.30 CScore: 0.30 Ok: 1,3,4,6,7,9,10,11,15,16,19 Err: 5,13,20 CompPred Top-5 - Sub: 7 Pred: 7 (1.00) Err: 0 (0.00) NErr: 0.00 WErr: 0.00 NWErr: 0.32 CScore: 0.32 Ok: 1,2,3,7,11,15,16 Err: CompPred Top-6 - Sub: 1 Pred: 1 (1.00) Err: 0 (0.00) NErr: 0.00 WErr: 0.00 NWErr: 0.48 CScore: 0.48 Ok: 16 Err: Randomization Summary ... Nullhypothesis: computed predictor can predict random classification with error rate <= x Composite Predictor Best - Tests: 100 - p(e <= 0.05) = 0.00 p(e <= 0.10) = 0.00 p(e <= 0.15) = 0.00 p(e <= 0.20) = 0.01 p(e <= 0.25) = 0.03 p(e <= 0.30) = 0.05 p(e <= 0.35) = 0.13 p(e <= 0.40) = 0.24 p(e <= 0.45) = 0.43 p(e <= 0.50) = 0.55 Computing top composite predictors at time 66.0 (without crossvalidation) ... Computing predictors at time 66.0 ... Defining composite predictors ... 1. Pred 0.4 - Pat: Pos - Sub: 19 Pred: 8 (0.42) Err: 0 (0.00) NErr: 0.29 WErr: 0.00 NWErr: 0.29 Ok: 2,3,6,8,12,15,17,19 Err: 2. Pred 11.1 - Pat: Neg - Sub: 19 Pred: 10 (0.53) Err: 1 (0.10) NErr: 0.29 WErr: 0.10 NWErr: 0.29 Ok: 1,2,3,6,9,12,15,16,19 Err: 20 3. Pred 0.30 - Pat: Neg - Sub: 19 Pred: 10 (0.53) Err: 1 (0.10) NErr: 0.29 WErr: 0.11 NWErr: 0.30 Ok: 3,4,6,7,8,11,15,16,19 Err: 13 4. Pred 8.0 - Pat: Pos - Sub: 19 Pred: 11 (0.58) Err: 2 (0.18) NErr: 0.32 WErr: 0.19 NWErr: 0.32 Ok: 2,3,7,10,11,15,16,18,19 Err: 5,13 Testing top composite predictors at time 66.0 ... Best Composite Predictor: Top-4 Score: 0.16 11.1[11.0] { CENTG2 ZBTB16 DLG5 } 0.4[0.0] { CNPY2 CETN2 PARK7 CSTA SEC11A METTL5 CHMP4A DNAJC15 H2AFZ ANXA2 ILF2 NDUFB8 NEDD8 CLEC4A POMP MRPS17 UFC1 PDHB ATP5J PPIB PSMD4 RPL36AL RPS6 SNRPD2 TTC1 TMEM14B DPM1 GADD45GIP1 RBM39 } 0.30 { DIDO1 NISCH VAMP1 ZNF767 } 8.0 { IFI30 NAGK }