Skip to main content
. 2021 Jan 7;7:626363. doi: 10.3389/fmolb.2020.626363

Table 3.

Performance of our deep learning-based method for predicting solvent exposure from protein sequence.

Scoring index Dataset
Cross-validation Blind test HVAR3D-2.0
MCC 0.63 0.63 0.60
Q2 (accuracy) 81% 82% 80%
F1 81% 82% 80%