Details
Original language | English |
---|---|
Pages (from-to) | 4825-4839 |
Number of pages | 15 |
Journal | Computational and structural biotechnology journal |
Volume | 19 |
Early online date | 18 Aug 2021 |
Publication status | Published - 2021 |
Abstract
Prediction of protein localization plays an important role in understanding protein function and mechanisms. In this paper, we propose a general deep learning-based localization prediction framework, MULocDeep, which can predict multiple localizations of a protein at both subcellular and suborganellar levels. We collected a dataset with 44 suborganellar localization annotations in 10 major subcellular compartments-the most comprehensive suborganelle localization dataset to date. We also experimentally generated an independent dataset of mitochondrial proteins in Arabidopsis thaliana cell cultures, Solanum tuberosum tubers, and Vicia faba roots and made this dataset publicly available. Evaluations using the above datasets show that overall, MULocDeep outperforms other major methods at both subcellular and suborganellar levels. Furthermore, MULocDeep assesses each amino acid's contribution to localization, which provides insights into the mechanism of protein sorting and localization motifs. A web server can be accessed at http://mu-loc.org.
Keywords
- Deep learning, Experimental benchmark datasets, Mechanism study, Protein localization, Web server
ASJC Scopus subject areas
- Biochemistry, Genetics and Molecular Biology(all)
- Genetics
- Biochemistry, Genetics and Molecular Biology(all)
- Biophysics
- Biochemistry, Genetics and Molecular Biology(all)
- Structural Biology
- Biochemistry, Genetics and Molecular Biology(all)
- Biochemistry
- Biochemistry, Genetics and Molecular Biology(all)
- Biotechnology
- Computer Science(all)
- Computer Science Applications
Cite this
- Standard
- Harvard
- Apa
- Vancouver
- BibTeX
- RIS
In: Computational and structural biotechnology journal, Vol. 19, 2021, p. 4825-4839.
Research output: Contribution to journal › Article › Research › peer review
}
TY - JOUR
T1 - MULocDeep
T2 - A deep-learning framework for protein subcellular and suborganellar localization prediction with residue-level interpretation
AU - Jiang, Yuexu
AU - Wang, Duolin
AU - Yao, Yifu
AU - Eubel, Holger
AU - Künzler, Patrick
AU - Møller, Ian Max
AU - Xu, Dong
N1 - Funding Information: This work was supported by the US National Institutes of Health grants R21-LM012790 and R35-GM126985. We would like to thank Dr. Hao Lin for providing suggestions in defining subcellular and suborganellar categories, and the anonymous reviewers for the helpful advice. We would like to thank Dr. Ning Zhang for providing the evaluation results by the MU-LOC method. This work used the high-performance computing infrastructure provided by Research Computing Support Services at the University of Missouri, as well as the Extreme Science and Engineering Discovery Environment (XSEDE), which is supported by National Science Foundation grant number ACI-1548562.
PY - 2021
Y1 - 2021
N2 - Prediction of protein localization plays an important role in understanding protein function and mechanisms. In this paper, we propose a general deep learning-based localization prediction framework, MULocDeep, which can predict multiple localizations of a protein at both subcellular and suborganellar levels. We collected a dataset with 44 suborganellar localization annotations in 10 major subcellular compartments-the most comprehensive suborganelle localization dataset to date. We also experimentally generated an independent dataset of mitochondrial proteins in Arabidopsis thaliana cell cultures, Solanum tuberosum tubers, and Vicia faba roots and made this dataset publicly available. Evaluations using the above datasets show that overall, MULocDeep outperforms other major methods at both subcellular and suborganellar levels. Furthermore, MULocDeep assesses each amino acid's contribution to localization, which provides insights into the mechanism of protein sorting and localization motifs. A web server can be accessed at http://mu-loc.org.
AB - Prediction of protein localization plays an important role in understanding protein function and mechanisms. In this paper, we propose a general deep learning-based localization prediction framework, MULocDeep, which can predict multiple localizations of a protein at both subcellular and suborganellar levels. We collected a dataset with 44 suborganellar localization annotations in 10 major subcellular compartments-the most comprehensive suborganelle localization dataset to date. We also experimentally generated an independent dataset of mitochondrial proteins in Arabidopsis thaliana cell cultures, Solanum tuberosum tubers, and Vicia faba roots and made this dataset publicly available. Evaluations using the above datasets show that overall, MULocDeep outperforms other major methods at both subcellular and suborganellar levels. Furthermore, MULocDeep assesses each amino acid's contribution to localization, which provides insights into the mechanism of protein sorting and localization motifs. A web server can be accessed at http://mu-loc.org.
KW - Deep learning
KW - Experimental benchmark datasets
KW - Mechanism study
KW - Protein localization
KW - Web server
UR - http://www.scopus.com/inward/record.url?scp=85114129650&partnerID=8YFLogxK
U2 - 10.1016/j.csbj.2021.08.027
DO - 10.1016/j.csbj.2021.08.027
M3 - Article
C2 - 34522290
VL - 19
SP - 4825
EP - 4839
JO - Computational and structural biotechnology journal
JF - Computational and structural biotechnology journal
SN - 2001-0370
ER -