Skip to main content
. 2021 Jan 14;11:1467. doi: 10.1038/s41598-021-81063-4

Figure 7.

Figure 7

LOGOCV scheme used for the comparison of the final Random Forest (RF) model with BLASTp. An RF model is iteratively trained on all but one of p held-out groups. In every iteration, the trained model is used to make predictions for the held-out group. The sequences in the same held-out group are also subjected to a local BLASTp search against the database without the held-out group. Finally, both results are compared.