Constant time approximation scheme for largest well predicted subset

The largest well predicted subset problem is formulated for comparison of two predicted 3D protein structures from the same sequence. A 3D protein structure is represented by an ordered point set A={a ₁,…,a _n }, where each a _i is a point in 3D space. Given two ordered point sets A={a ₁,…,a _n } and B={b ₁,b ₂,…b _n } containing n points, and a threshold d, the largest well predicted subset problem is to find the rigid body transformation T for a largest subset B _opt of B such that the distance between a _i and T(b _i ) is at most d for every b _i in B _opt . A meaningful prediction requires that the size of B _opt is at least αn for some constant α (Li et al. in CPM 2008, 2008). We use LWPS(A,B,d,α) to denote the largest well predicted subset problem with meaningful prediction. An (1+δ ₁,1?δ ₂)-approximation for LWPS(A,B,d,α) is to find a transformation T to bring a subset B′?B of size at least (1?δ ₂)|B _opt | such that for each b _i ∈B′, the Euclidean distance between the two points distance?(a _i ,T(b _i ))≤(1+δ ₁)d. We develop a constant time (1+δ ₁,1?δ ₂)-approximation algorithm for LWPS(A,B,d,α) for arbitrary positive constants δ ₁ and δ ₂. To our knowledge, this is the first constant time algorithm in this area. Li et al. (CPM 2008, 2008) showed an $O(n(\log n)^{2}/\delta_{1}^{5})$ time randomized (1+δ ₁)-distance approximation algorithm for the largest well predicted subset problem under meaningful prediction. We also study a closely related problem, the bottleneck distance problem, where we are given two ordered point sets A={a ₁,…,a _n } and B={b ₁,b ₂,…b _n } containing n points and the problem is to find the smallest d _opt such that there exists a rigid transformation T with distance(a _i ,T(b _i ))≤d _opt for every point b _i ∈B. A (1+δ)-approximation for the bottleneck distance problem is to find a transformation T, such that for each b _i ∈B, distance?(a _i ,T(b _i ))≤(1+δ)d _opt , where δ is a constant. For an arbitrary constant δ, we obtain a linear O(n/δ ⁶) time (1+δ)-algorithm for the bottleneck distance problem. The best known algorithms for both problems require super-linear time (Li et al. in CPM 2008, 2008).