NLScore: a novel quantitative algorithm based on 3 dimensional structural determinants to predict the probability of nuclear localization in proteins containing classical nuclear localization signals

NLScore: a novel quantitative algorithm based on 3 dimensional structural determinants to predict... The presence of a nuclear localization signal (NLS) in proteins can be inferred by the presence of a stretch of basic amino acids (KRKK). These NLSs are termed classical NLS (cNLS). However, only a fraction of proteins containing the cNLS pattern are transported into the nucleus by binding to importin α. Hence, there must exist, additional structural determinants that guide the appropriate interaction between putative NLSs containing cargo and importin α. Using 52 protein structures containing cNLS obtained from RCSB PDB, we assembled a training set and a validation set such that both sets were comprised of a combination of proteins with proven nuclear localization and ones that were non-nuclear. We modeled the interface between cargoes containing cNLS and importin α. We conducted rigid body docking and produced induced-fit modes by allowing both side chain and the backbone to be flexible. The output of these studies and additional determinants such as energy of interaction, atomic contacts, hydrophilic interaction, cationic interaction, and penetration of the cargo protein were used to derive a 26 parameter quantitative structure activity relationship based regression equation. This was further optimized by a step-wise backward elimination approach to derive a 15 parameter score. This NLScore was not only able to correctly classify confirmed nuclear and non-nuclear localized proteins but it was able to perform better than currently implemented algorithms like NucPred, Euk-mPLoc 2.0, cNls Mapper, and NLStradamus. Leave-one-out cross validation (LOOCV) showed that NLScore correctly predicted 78.6% and 81.6% of non-nuclear and nuclear proteins respectively. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Molecular Modeling Springer Journals

NLScore: a novel quantitative algorithm based on 3 dimensional structural determinants to predict the probability of nuclear localization in proteins containing classical nuclear localization signals

Loading next page...
 
/lp/springer_journal/nlscore-a-novel-quantitative-algorithm-based-on-3-dimensional-b7hhy8Tgbf
Publisher
Springer Journals
Copyright
Copyright © 2017 by Springer-Verlag GmbH Germany
Subject
Chemistry; Computer Applications in Chemistry; Molecular Medicine; Computer Appl. in Life Sciences; Characterization and Evaluation of Materials; Theoretical and Computational Chemistry
ISSN
1610-2940
eISSN
0948-5023
D.O.I.
10.1007/s00894-017-3420-y
Publisher site
See Article on Publisher Site

Abstract

The presence of a nuclear localization signal (NLS) in proteins can be inferred by the presence of a stretch of basic amino acids (KRKK). These NLSs are termed classical NLS (cNLS). However, only a fraction of proteins containing the cNLS pattern are transported into the nucleus by binding to importin α. Hence, there must exist, additional structural determinants that guide the appropriate interaction between putative NLSs containing cargo and importin α. Using 52 protein structures containing cNLS obtained from RCSB PDB, we assembled a training set and a validation set such that both sets were comprised of a combination of proteins with proven nuclear localization and ones that were non-nuclear. We modeled the interface between cargoes containing cNLS and importin α. We conducted rigid body docking and produced induced-fit modes by allowing both side chain and the backbone to be flexible. The output of these studies and additional determinants such as energy of interaction, atomic contacts, hydrophilic interaction, cationic interaction, and penetration of the cargo protein were used to derive a 26 parameter quantitative structure activity relationship based regression equation. This was further optimized by a step-wise backward elimination approach to derive a 15 parameter score. This NLScore was not only able to correctly classify confirmed nuclear and non-nuclear localized proteins but it was able to perform better than currently implemented algorithms like NucPred, Euk-mPLoc 2.0, cNls Mapper, and NLStradamus. Leave-one-out cross validation (LOOCV) showed that NLScore correctly predicted 78.6% and 81.6% of non-nuclear and nuclear proteins respectively.

Journal

Journal of Molecular ModelingSpringer Journals

Published: Aug 9, 2017

References

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off