Table 1.

List of all features considered.

Features	Description
Nucleotide level
Microdeletion/microinsertion positions (2)	Distances to nearest 5' and 3' splicing positions
DNA conservation scores (3)	Maximum, minimum, average
Protein level
Evolution feature (30)	Maximum, minimum, average values (7 transition probabilities between match (M), microdeletion (D), and microinsertion (I) (MM, MI, MD, IM, II, DM, DD), 3 effective numbers of match/microinsertion/microdeletion)
Length (4)	Protein length, Microdeletion/microinsertion length, Distances to terminals
ΔS (1)	The indel-induced change to the HMM match score
Disorder score (3)	Maximum, minimum, average
Secondary structure (12)	Maximum, minimum, average probability (C, H, E), Predicted secondary structure (C, H, E)
Accessible surface area (3)	Maximum, minimum, average