Skip to main content
. 2021 Mar 3;12:562199. doi: 10.3389/fmicb.2021.562199

Figure 9.

Figure 9

Scanned image of a DE output of positive and negative controls of MilBVV (A) and reconstituted image (B) made by pixels with predicted probabilities above the 0.135 cutoff based on the LBVaV-trained model. Lanes A and B show duplicates of undiluted (1 and 5), and 1:5 (2) 1:10 (3), and 1:50 (4) dilutions of the positive control of MilBVV. Dots 6 and 7 are buffer only controls. Dots 8 to 12 contain the negative control of MilBVV undiluted (8, 12), and at 1:5 (9), 1:10 (10), and 1:50 (11) dilutions. The translation of unique antigens, of MilBVV here, to the universal basic color language allows for its detection by the algorithm (model) we built, even though it was trained on the basis of a different virus, LBVaV.