Smooth orientation-dependent scoring function (SBROD) for coarse-grained protein quality assessment uses only the conformation of the protein backbone, and hence it can be applied to scoring the coarse-grained protein models.
The workflow of SBROD consists in two stages. First, the method extracts features from each protein model in the dataset. Then, the scoring function assigns a score to each processed protein model depending on its features extracted at the first stage. Figure above schematically shows the workflow of SBROD. Here, four types of inter-atomic interactions, described in details below, are taken into account when extracting the features. After these features have been extracted and preprocessed, a Ridge Regression model is trained on them to predict the GDT-TS of protein models.
Mikhail Karasikov1-3, Guillaume Pagès3 & Sergei Grudinin3,1Center for Energy Systems, Skolkovo Institute of Science and Technology, Moscow, 143026, Russia 2Moscow Institute of Physics and Technology, Moscow, 141701, Russia
3Nano-D team, Inria/CNRS Grenoble, Francee-mail: Sergei.Grudinin @ inria.fr
Binaries, source code and tutorials can be found at https://gitlab.inria.fr/grudinin/sbrod
The server is running at www.karasikov.com/proteins/
SBROD was also evaluated in CASP13 as a QA server.
- Mikhail Karasikov, Guillaume Pagès & Sergei Grudinin, “Smooth orientation-dependent scoring function for coarse-grained protein quality assessment”. Bioinformatics, bty1037, https://doi.org/10.1093/bioinformatics/bty1037
All rights reserved. The academic version is free.