Ranking predicted protein structures with support vector regression

Abstract

Protein structure prediction is an important problem of both intellectual and practical interest. Most protein structure prediction approaches generate multiple candidate models first, and then use a scoring function to select the best model among these candidates. In this work, we develop a scoring function using support vector regression (SVR). Both consensus-based features and features from individual structures are extracted from a training data set containing native protein structures and predicted structural models submitted to CASP5 and CASP6. The SVR learns a scoring function that is a linear combination of these features. We test this scoring function on two data sets. First, when used to rank server models submitted to CASP7, the SVR score selects predictions that are comparable to the best performing server in CASP7, Zhang-Server, and significantly better than all the other servers. Even if the SVR score is not allowed to select Zhang-Server models, the SVR score still selects predictions that are significantly better than all the other servers. In addition, the SVR is able to select significantly better models and yield significantly better Pearson correlation coefficients than the two best Quality Assessment groups in CASP7, QA556 (LEE), and QA634 (Pcons). Second, this work aims to improve the ability of the Robetta server to select best models, and hence we evaluate the performance of the SVR score on ranking the Robetta server template-based models for the CASP7 targets. The SVR selects significantly better models than the Robetta K*Sync consensus alignment score.

On-line hierarchy of general linear models for selecting and ranking the best predicted protein structures

“…3dSim is the average of the MaxSub scores between a structure and the other structures predicted for the same target protein provided that the MaxSub similarity score between the two models is greater than 0.4. Similar scores are implemented by [4], [5], [7], [21]. Equations 1 and 2 show how the 3dSim score is computed.…”

“…In other words, we standardize the values of each score with regard to the structures predicted to the same target protein only. Our target-wise normalization is different from the method implemented in [7]. Equation 3 shows how to normalize a score value v of a structure i which is predicted to target protein t.…”

“…Related model quality assessment programs [3]- [7] are based on machine learning algorithm. The learning algorithm learns a function that maps a combination of a set of features extracted form the computationally generated structures to a quality score such as the RMSD score [8], the MaxSub score [9], the TM score [10], and the GDT score [11].…”

“…The developers of these methods formulated the problem as a regression problem. The majority of these methods except the one developed by Qiu et al [7] treat all structures the same way during training, while Qiu's method pays attention to the high quality structures during training. In our work we also distinguish between the high quality structures from the low quality structures during training and testing.…”

An iterative self‐refining and self‐evaluating approach for protein model quality estimation

“…[4][5][6][7][8][9][10] Single-model methods directly predict the quality of a model from its structural features using machine learning, statistical, or physical methods. [11][12][13][14][15][16]21,22 According to recent CASP experiments, 17 multiple-model methods are currently more accurate than single-model methods, although they do not work well if only a small number of models are available or the structures of input models are largely different. Another drawback is that clustering methods usually need relatively long computational time that makes it less efficient and less feasible to be used in daily research.…”