Explaining early folding through interpretable machine learning
This page shows summarised information on the rules the SlGb predictor uses to predict
whether an amino acid fold early given its local sequence context. The data shown is
Default: PISCES set of 3020 proteins
Scrambled: Default set, scrambled
Denovo: 98 de novo designed proteins
dataset, you can change dataset by changing the selection. To see the data for a specific fragment
of the currently selected dataset for a specific prediction method, use the selection on the left-hand side
and press 'Submit'.
Click for information on below plot
This diagram shows the key features that the machine learning identified for the prediction for the
complete default dataset by the SlGb method.