Skip to main content

Table 2 Coefficients of linear discriminant functions (LDF) derived from LDA for all possible combination of antibodies

From: Discriminant analysis and machine learning approach for evaluating and improving the performance of immunohistochemical algorithms for COO classification of DLBCL

Antibody combination Sens Spec COO Constant Antibody
1 2 3 4 5
CD10 BCL6 FOXP1 GCET1 MUM1
As in Nyman 3,5 0.81 0.74 GCB − 0.57    2.47   1.11
Non-GCB − 3.29    4.38   5.06
As in Hans and 1,2,5 0.86 0.88 GCB − 4.21 5.01 7.00    1.05
As in Colomo     Non-GCB − 3.09 0.20 4.99    5.69
As in Hans* 1,5 0.77 0.91 GCB − 2.29 6.53     2.11
Non-GCB − 2.12 1.28     6.45
As in Choi 1,2,3,4,5 0.87 0.91 GCB − 4.80 4.59 6.31 0.71 3.28 1.02
Non-GCB − 3.98 − 0.41 3.80 3.75 1.24 4.73
As in Choi* 1,3,4,5 0.86 0.91 GCB − 3.34 5.67   1.78 4.03 1.67
Non-GCB − 3.46 0.24   4.39 1.69 5.12
As in VY3 1,2,3 0.87 0.91 GCB − 4.18 4.86 6.99 0.64   
Non-GCB − 2.90 − 0.74 4.58 4.61   
As in VY4 1,2,3,4 0.84 0.90 GCB − 4.75 4.49 6.44 0.91 3.26  
Non-GCB − 2.97 − 0.86 4.40 4.70 1.12  
1,4,5 0.81 0.92 GCB − 3.14 6.01    3.93 2.22
Non-GCB − 2.23 1.09    1.44 6.49
  1. Columns (Antibody) show the coefficient associated with each antibody for GCB classification (Sensibility performance) and non-GCB classification (Specificity performance). The first column remarks the combinations of antibodies as were used by IHC-decision trees. BCL6 and CD10 in the LDF are strongly related to the classification of GCB cases, whereas the inclusion of MUM1 in any algorithm is related to non-GCB detection, getting the most considerable value in the LDF
  2. Numeric Tags 1 = CD10, 2 = BCL6, 3 = FOXP1, 4 = GCTE1, and 5 = MUM1