551) Epigenetic inhibitor library ( Table 3). As shown in many studies, total IgE values did not correlate with ImmunoCAP results ( Table 3) and were also

unable to discriminate between children who acquired tolerance and children who were still sensitive to milk up their last visit (p = 0.305 ANOVA). ImmunoCAP values for Cow’s milk, Casein, β-lactoglobulin (p = < 0.001) but not α-lactalbumin (p = 0.401) were able to make this discrimination. Furthermore, within the cohort that acquired milk tolerance during the time span of these visits, there was a small but direct correlation of ImmunoCAP values and age of tolerance i.e., higher casein or total cow's milk ImmunoCAP values in children that acquired milk tolerance at a later age ( Table 3). These results are in agreement with the larger specific average

IgE values shown by the susceptible group in the array data summary presented in Fig. 2. A cross-validated Partial Least Squares Regression (PLSR) model was generated between the array data and the ImmunoCAP results and shown in Fig. 5. The best PLSR fit was achieved with Casein ImmunoCAP values (model fit R2 = 0.7; cross validation R2 = 0.6) but regression was less efficient for cow’s milk (R2 = 0.57 and 0.45 for model and cross validation respectively). Both models showed strongest predictive contributions from dairy proteins as expected and shown in Fig. 5B. PLS-DA models that directly predicted onset of tolerance based only on IgE array data did not result in accurate models, only predicting 2/3 of the tolerant cases correctly. Whether the rate of variation of the specific IgE content with successive visits had a better predictive selleck chemical power was investigated using the overall cumulative variation and the variation of each patient per year (Fig. 6). Overall the responses were very homogeneous

with some exceptions. One patient for instance has shown an increase in specific Amino acid IgE values with most of the groups tested. This contrasts with another patient showing an increase in the specific IgE response to dairy products only. Most of the remaining patients showed a diminishing dairy IgE response with time (Fig. 6). The slope of variation with time, variance and covariance of the measurements were not significantly predictive of any of the clinical parameters analyzed. Conversely, corroborating the data described earlier between ImmunoCAP Casein and the age of onset of milk tolerance (Table 3), the regression analysis of the specific IgE array data employing partial least square method (PLS) was also able to establish a relevant cross validated fit (R2 = 0.695) for this variable ( Fig. 7). These coefficients were obtained when the products were clustered in groups as variables. A higher cross-validation coefficient (R2 = 0.701) was obtained using the individual measurement values instead of clustered groups (not shown), however, the interpretation becomes more cumbersome due to the amount of variables involved.