Zoomqa: Residue-Level Single-Model QA Support Vector Machine Utilizing Sequential and 3D Structural Features

Abstract

MotivationThe Estimation of Model Accuracy problem is a cornerstone problem in the field of Bioinformatics. When predictions are made for proteins of which we do not know the native structure, we run into an issue to tell how good a tertiary structure prediction is, especially the protein binding regions, which are useful for drug discovery. Currently, most methods only evaluate the overall quality of a protein decoy, and few can work on residue level and protein complex. Here we introduce ZoomQA, a novel, single-model method for assessing the accuracy of a tertiary protein structure / complex prediction at residue level. ZoomQA differs from others by considering the change in chemical and physical features of a fragment structure (a portion of a protein within a radius r of the target amino acid) as the radius of contact increases. Fourteen physical and chemical properties of amino acids are used to build a comprehensive representation of every residue within a protein and grades their placement within the protein as a whole. Moreover, ZoomQA can evaluate the quality of protein complex, which is unique.ResultsWe benchmark ZoomQA on CASP14, it outperforms other state of the art local QA methods and rivals state of the art QA methods in global prediction metrics. Our experiment shows the efficacy of these new features, and shows our method is able to match the performance of other state-of-the-art methods without the use of homology searching against database or PSSM matrix.Availabilityhttp://zoomQA.renzhitech.comContactcaora@plu.edu

Estimation of Model Accuracy By A Unique Set of Features And Tree-Based Regressor

“…This achievement would inevitably have a profound impact on the field of protein structure prediction, including the accuracy estimation sub-task. We conclude this manuscript with some speculations regarding the future role of accuracy estimation in a new era of accurate protein structure prediction.Most recent EMA methods use machine learning (ML) algorithms, including neural networks [22][23][24] , SVM [25][26][27] , and treebased models 28 , to create a statistical model that combines measurable features into a single number, which estimates decoy quality 25,29 . To this end, ML algorithms use datasets of annotated decoys and learn the intricate relations between the features and decoy quality.…”

“…Most recent EMA methods use machine learning (ML) algorithms, including neural networks [22][23][24] , SVM [25][26][27] , and treebased models 28 , to create a statistical model that combines measurable features into a single number, which estimates decoy quality 25,29 . To this end, ML algorithms use datasets of annotated decoys and learn the intricate relations between the features and decoy quality.…”