Algorithm 2.

A New Selective Consensus Method, SelCon

Require: Predicted protein models S = {si, 1 ≤ i ≤ p} and a weight function W.

For each si ∈ S

$$x_i=\frac{{\displaystyle \sum _{s_j\in S}}W(\mathit{sim}(s_i,s_j))·\mathit{sim}(s_i,s_j)}{{\displaystyle \sum _{s_j\in S}}W(\mathit{sim}(s_i,s_j))}$$ (5)

return QA score for each structure, X = {xi, 1 ≤ i ≤ p}.