MS-<it>k</it>NN: protein function prediction by integrating multiple data sources
<p>Abstract</p> <p>Background</p> <p>Protein function determination is a key challenge in the post-genomic era. Experimental determination of protein functions is accurate, but time-consuming and resource-intensive. A cost-effective alternative is to use the known infor...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
BMC
2013-02-01
|
| Series: | BMC Bioinformatics |
| _version_ | 1830349657795985408 |
|---|---|
| author | Lan Liang Djuric Nemanja Guo Yuhong Vucetic Slobodan |
| author_facet | Lan Liang Djuric Nemanja Guo Yuhong Vucetic Slobodan |
| author_sort | Lan Liang |
| collection | DOAJ |
| description | <p>Abstract</p> <p>Background</p> <p>Protein function determination is a key challenge in the post-genomic era. Experimental determination of protein functions is accurate, but time-consuming and resource-intensive. A cost-effective alternative is to use the known information about sequence, structure, and functional properties of genes and proteins to predict functions using statistical methods. In this paper, we describe the Multi-Source <it>k</it>-Nearest Neighbor (MS-<it>k</it>NN) algorithm for function prediction, which finds <it>k</it>-nearest neighbors of a query protein based on different types of similarity measures and predicts its function by weighted averaging of its neighbors' functions. Specifically, we used 3 data sources to calculate the similarity scores: sequence similarity, protein-protein interactions, and gene expressions.</p> <p>Results</p> <p>We report the results in the context of 2011 Critical Assessment of Function Annotation (CAFA). Prior to CAFA submission deadline, we evaluated our algorithm on 1,302 human test proteins that were represented in all 3 data sources. Using only the sequence similarity information, MS-<it>k</it>NN had term-based Area Under the Curve (AUC) accuracy of Gene Ontology (GO) molecular function predictions of 0.728 when 7,412 human training proteins were used, and 0.819 when 35,622 training proteins from multiple eukaryotic and prokaryotic organisms were used. By aggregating predictions from all three sources, the AUC was further improved to 0.848. Similar result was observed on prediction of GO biological processes. Testing on 595 proteins that were annotated after the CAFA submission deadline showed that overall MS-<it>k</it>NN accuracy was higher than that of baseline algorithms Gotcha and BLAST, which were based solely on sequence similarity information. Since only 10 of the 595 proteins were represented by all 3 data sources, and 66 by two data sources, the difference between 3-source and one-source MS-<it>k</it>NN was rather small.</p> <p>Conclusions</p> <p>Based on our results, we have several useful insights: (1) the <it>k</it>-nearest neighbor algorithm is an efficient and effective model for protein function prediction; (2) it is beneficial to transfer functions across a wide range of organisms; (3) it is helpful to integrate multiple sources of protein information.</p> |
| first_indexed | 2024-12-20T00:07:46Z |
| format | Article |
| id | doaj.art-ed443511b3694526b1a0cd91dd1d7a11 |
| institution | Directory Open Access Journal |
| issn | 1471-2105 |
| language | English |
| last_indexed | 2024-12-20T00:07:46Z |
| publishDate | 2013-02-01 |
| publisher | BMC |
| record_format | Article |
| series | BMC Bioinformatics |
| spelling | doaj.art-ed443511b3694526b1a0cd91dd1d7a112022-12-21T20:00:36ZengBMCBMC Bioinformatics1471-21052013-02-0114Suppl 3S810.1186/1471-2105-14-S3-S8MS-<it>k</it>NN: protein function prediction by integrating multiple data sourcesLan LiangDjuric NemanjaGuo YuhongVucetic Slobodan<p>Abstract</p> <p>Background</p> <p>Protein function determination is a key challenge in the post-genomic era. Experimental determination of protein functions is accurate, but time-consuming and resource-intensive. A cost-effective alternative is to use the known information about sequence, structure, and functional properties of genes and proteins to predict functions using statistical methods. In this paper, we describe the Multi-Source <it>k</it>-Nearest Neighbor (MS-<it>k</it>NN) algorithm for function prediction, which finds <it>k</it>-nearest neighbors of a query protein based on different types of similarity measures and predicts its function by weighted averaging of its neighbors' functions. Specifically, we used 3 data sources to calculate the similarity scores: sequence similarity, protein-protein interactions, and gene expressions.</p> <p>Results</p> <p>We report the results in the context of 2011 Critical Assessment of Function Annotation (CAFA). Prior to CAFA submission deadline, we evaluated our algorithm on 1,302 human test proteins that were represented in all 3 data sources. Using only the sequence similarity information, MS-<it>k</it>NN had term-based Area Under the Curve (AUC) accuracy of Gene Ontology (GO) molecular function predictions of 0.728 when 7,412 human training proteins were used, and 0.819 when 35,622 training proteins from multiple eukaryotic and prokaryotic organisms were used. By aggregating predictions from all three sources, the AUC was further improved to 0.848. Similar result was observed on prediction of GO biological processes. Testing on 595 proteins that were annotated after the CAFA submission deadline showed that overall MS-<it>k</it>NN accuracy was higher than that of baseline algorithms Gotcha and BLAST, which were based solely on sequence similarity information. Since only 10 of the 595 proteins were represented by all 3 data sources, and 66 by two data sources, the difference between 3-source and one-source MS-<it>k</it>NN was rather small.</p> <p>Conclusions</p> <p>Based on our results, we have several useful insights: (1) the <it>k</it>-nearest neighbor algorithm is an efficient and effective model for protein function prediction; (2) it is beneficial to transfer functions across a wide range of organisms; (3) it is helpful to integrate multiple sources of protein information.</p> |
| spellingShingle | Lan Liang Djuric Nemanja Guo Yuhong Vucetic Slobodan MS-<it>k</it>NN: protein function prediction by integrating multiple data sources BMC Bioinformatics |
| title | MS-<it>k</it>NN: protein function prediction by integrating multiple data sources |
| title_full | MS-<it>k</it>NN: protein function prediction by integrating multiple data sources |
| title_fullStr | MS-<it>k</it>NN: protein function prediction by integrating multiple data sources |
| title_full_unstemmed | MS-<it>k</it>NN: protein function prediction by integrating multiple data sources |
| title_short | MS-<it>k</it>NN: protein function prediction by integrating multiple data sources |
| title_sort | ms it k it nn protein function prediction by integrating multiple data sources |
| work_keys_str_mv | AT lanliang msitkitnnproteinfunctionpredictionbyintegratingmultipledatasources AT djuricnemanja msitkitnnproteinfunctionpredictionbyintegratingmultipledatasources AT guoyuhong msitkitnnproteinfunctionpredictionbyintegratingmultipledatasources AT vuceticslobodan msitkitnnproteinfunctionpredictionbyintegratingmultipledatasources |