Table 3.
List of selected features for different training sets
| Deletions | Insertions | indels | Non-redundant indels |
|---|---|---|---|
| Disorder(min) | Disorder(min) | Disorder(min) | Disorder(min) |
| DNA conservation(max) | DNA conservation(max) | DNA conservation(max) | DNA conservation(max) |
| Deletion length | P(m-i)e (min) | ΔSd | ΔSd |
| ASAa (min) | ΔSd | Neffc(ave) | Neffc(min) |
| P(m-d)b(ave) | P(m-i)e (ave) | indel length | ASAa (ave) |
| Neffc(min) | Disorder(ave) | Distance to the nearest splicing site (upstream) | indel length |
| Distance to the nearest splicing site (downstream) | Helical probability(max) | ASAa (max) | ASAa (max) |
| ASAa(max) | P(m-m)f(ave) | Neffc(min) | P(m-m)f(max) |
| ΔSd | DNA conservation(ave) | ||
| ASA(ave) |
aASA, solvent accessible surface area. bP(m-d), match-to-deletion transition probability. cNeff: the number of effective homologous sequences aligned to residues. dΔS, indel-induced change to alignment score. eP(m-i), match-to-insertion transition probability. fP(m-m), match-to-match transition probability.