Positive-unlabeled learning for disease gene identification
Open accessThe DOI matches its Crossref record; 59 of the 59 references it cites were found in a registry or an index.
Abstract
Background: Identifying disease genes from human genome is an important but challenging task in biomedical research. Machine learning methods can be applied to discover new disease genes based on the known ones. Existing machine learning methods typically use the known disease genes as the positive training set P and the unknown genes as the negative training set N (non-disease gene set does not exist) to build classifiers to identify new disease genes from the unknown genes. However, such kind of classifiers is actually built from a noisy negative set N as there can be unknown disease genes in N itself. As a result, the classifiers do not perform as well as they could be. Result: Instead of treating the unknown genes as negative examples in N, we treat them as an unlabeled set U. We design a novel positive-unlabeled (PU) learning algorithm PUDI (PU learning for disease gene identification) to build a classifier using P and U. We first partition U into four sets, namely, reliable negative set RN, likely positive set LP, likely negative set LN and weak negative set WN. The weighted support vector machines are then used to build a multi-level classifier based on the four training sets and positive training set P to identify disease genes. Our experimental results demonstrate that our proposed PUDI algorithm outperformed the existing methods significantly. Conclusion: The proposed PUDI algorithm is able to identify disease genes more accurately by treating the unknown data more appropriately as unlabeled set U instead of negative set N. Given that many machine learning problems in biomedical research do involve positive and unlabeled data instead of negative data, it is possible that the machine learning methods for these problems can be further improved by adopting PU learning methods, as we have done here for disease gene identification. Availability and implementation: The executable program and data are available at http://www1.i2r.a-star.edu.sg/∼xlli/PUDI/PUDI.html. Contact: xlli@i2r.a-star.edu.sg or yang0293@e.ntu.edu.sg Supplementary information: Supplementary Data are available at Bioinformatics online.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Collective dynamics of ‘small-world’ networks
Watts, D. J., & Strogatz, S. H. (1998) · Nature · doi:10.1038/30918 · Cited by 43,060
LIBSVM: A library for support vector machines
Chang, C.-C., & Lin, C.-J. (2011) · ACM Transactions on Intelligent Systems and Technology · doi:10.1145/1961189.1961199 · Cited by 40,918 · cited 1×
“To take into account of the inherently different levels of trustworthiness of labels in P, RN, LP, LN and WN, we use a multi-level examples learning technique, weighted SVMs (Chang and Lin, 2011; Vapink, 1998), to build a classifier.”
MethodWu, Y. (1999) · Technometrics · doi:10.1080/00401706.1999.10485951 · Cited by 27,221 · cited 1×
“To take into account of the inherently different levels of trustworthiness of labels in P, RN, LP, LN and WN, we use a multi-level examples learning technique, weighted SVMs (Chang and Lin, 2011; Vapink, 1998), to build a classifier.”
MethodThe Pfam Protein Families Database
Bateman, A. (2002) · Nucleic Acids Research · doi:10.1093/nar/30.1.276 · Cited by 14,000
The Gene Ontology (GO) database and informatics resource
Gene Ontology Consortium (2004) · Nucleic Acids Research · doi:10.1093/nar/gkh036 · Cited by 4,579 · cited 1×
“A more biologically meaningful approach would be to build a binary classification model that can automatically classify a gene as disease or not.”
BackgroundA STATISTICAL INTERPRETATION OF TERM SPECIFICITY AND ITS APPLICATION IN RETRIEVAL
SPARCK JONES, K. (1972) · Journal of Documentation · doi:10.1108/eb026526 · Cited by 4,505
Goh, K.-I., Cusick, M. E., Valle, D., Childs, B., Vidal, M., & Barabási, A.-L. (2007) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.0701361104 · Cited by 3,242 · cited 1×
“However, such kind of classifiers is actually built from a noisy negative set N as there can be unknown disease genes in”
BackgroundHuman Protein Reference Database--2009 update
Keshava Prasad, T. S., Goel, R., Kandasamy, K., Keerthikumar, S., Kumar, S., Mathivanan, S., Telikicherla, D., Raju, R., Shafreen, B., Venugopal, A., et al. (2009) · Nucleic Acids Research · doi:10.1093/nar/gkn892 · Cited by 3,234
The Pfam protein families database
Finn, R. D., Mistry, J., Tate, J., Coggill, P., Heger, A., Pollington, J. E., Gavin, O. L., Gunasekaran, P., Ceric, G., Forslund, K., et al. (2010) · Nucleic Acids Research · doi:10.1093/nar/gkp985 · Cited by 2,694 · cited 2×
“Köhler et al. (2008) made use of the observation that proteins caused by same/similar disorders are likely attached together in protein–protein interaction (PPI) network (Gandhi et al., 2006) and applied the random walk algorithm on the PPI network for disease gene prioritization.”
MethodGeneCards Version 3: the human gene integrator
Safran, M., Dalah, I., Alexander, J., Rosen, N., Iny Stein, T., Shmoish, M., Nativ, N., Bahir, I., Doniger, T., Krug, H., et al. (2010) · Database · doi:10.1093/database/baq020 · Cited by 2,088
Random Walks on Graphs: A Survey
Lovász, L. (2001) · Cited by 1,411
A new method to measure the semantic similarity of GO terms
Wang, J. Z., Du, Z., Payattakool, R., Yu, P. S., & Chen, C.-F. (2007) · Bioinformatics · doi:10.1093/bioinformatics/btm087 · Cited by 1,377
Walking the Interactome for Prioritization of Candidate Disease Genes
Köhler, S., Bauer, S., Horn, D., & Robinson, P. N. (2008) · The American Journal of Human Genetics · doi:10.1016/j.ajhg.2008.02.013 · Cited by 1,258
Deza, M. M., & Deza, E. (2014) · Springer eBooks · doi:10.1007/978-3-662-44342-2 · Cited by 1,041
Deza, E., & Deza, M. M. (2009) · Springer eBooks · doi:10.1007/978-3-642-00234-2 · Cited by 950 · cited 4×
“For each gene gi in U, if its Euclidean distance dist(pr, Vgi)4Ave_dist, we regard it as a reliable negative example and store it in RN (Lines 7–9); since it is very far away from the positive examples, it is thus safe for us to treat it as a negative example.”
MethodInfluentialIdeker, T., & Sharan, R. (2008) · Genome Research · doi:10.1101/gr.071852.107 · Cited by 817 · cited 1×
“, 2008), high functional similarities (Ideker and Sharan, 2008) and physical interactions between their gene products (Brunner et al.”
BackgroundMendelian Inheritance in Man and Its Online Version, OMIM
McKusick, V. A. (2007) · The American Journal of Human Genetics · doi:10.1086/514346 · Cited by 739
Online Predicted Human Interaction Database
Brown, K. R., & Jurisica, I. (2005) · Bioinformatics · doi:10.1093/bioinformatics/bti273 · Cited by 606
Flicek, P., Amode, M. R., Barrell, D., Beal, K., Brent, S., Chen, Y., Clapham, P., Coates, G., Fairley, S., Fitzgerald, S., et al. (2011) · Nucleic Acids Research · doi:10.1093/nar/gkq1064 · Cited by 594
Partially Supervised Classification of Text Documents
Liu, B., Lee, W. S., Yu, P. S., & Li, X. (2002) · Cited by 534
Self-citationDiscovering disease-genes by topological features in human protein–protein interaction network
Xu, J., & Li, Y. (2006) · Bioinformatics · doi:10.1093/bioinformatics/btl467 · Cited by 461
Gandhi, T. K. B., Zhong, J., Mathivanan, S., Karthick, L., Chandrika, K. N., Mohan, S. S., Sharma, S., Pinkert, S., Nagaraju, S., Periaswamy, B., et al. (2006) · Nature Genetics · doi:10.1038/ng1747 · Cited by 442
Genome-wide inferring gene–phenotype relationship by walking on the heterogeneous network
Li, Y., & Patra, J. C. (2010) · Bioinformatics · doi:10.1093/bioinformatics/btq108 · Cited by 418
The modular nature of genetic diseases
Oti, M., & Brunner, H. (2007) · Clinical Genetics · doi:10.1111/j.1399-0004.2006.00708.x · Cited by 410
Association of genes to genetically inherited diseases using data mining
Perez-Iratxeta, C., Bork, P., & Andrade, M. A. (2002) · Nature Genetics · doi:10.1038/ng895 · Cited by 354
Genome-wide identification of genes likely to be involved in human genetic disease
Lopez-Bigas, N. (2004) · Nucleic Acids Research · doi:10.1093/nar/gkh605 · Cited by 261
Human Protein Reference Database
Wilson, N. (2004) · Nature Reviews Molecular Cell Biology · doi:10.1038/nrm1305 · Cited by 250
Speeding disease gene discovery by sequence based candidate prioritization
Adie, E. A., Adams, R. R., Evans, K. L., Porteous, D. J., & Pickard, B. S. (2005) · BMC Bioinformatics · doi:10.1186/1471-2105-6-55 · Cited by 249 · cited 1×
“Adie et al. (2005) further improved on this method by employing a decision tree algorithm based on a variety of genomic and evolutionary features, such as coding sequence length, evolutionary conservation, presence, closeness of paralogs in the human genome, etc.”
MethodIn-silico human genomics with GeneCards
Stelzer, G., Dalah, I., Stein, T., Satanower, Y., Rosen, N., Nativ, N., Oz-Levi, D., Olender, T., Belinky, F., Bahir, I., et al. (2011) · Human Genomics · doi:10.1186/1479-7364-5-6-709 · Cited by 236
Linghu, B., Snitkin, E. S., Hu, Z., Xia, Y., & DeLisi, C. (2009) · Genome Biology · doi:10.1186/gb-2009-10-9-r91 · Cited by 232
POCUS: mining genomic sequence annotation to predict disease genes
Turner, F. S., Clutterbuck, D. R., & Semple, C. A. (2003) · Genome Biology · doi:10.1186/gb-2003-4-11-r75 · Cited by 225
Martinez-Arca, S., Alberts, P., Zahraoui, A., Louvard, D., & Galli, T. (2000) · The Journal of Cell Biology · doi:10.1083/jcb.149.4.889 · Cited by 223
Lysophosphatidic Acid Receptors Determine Tumorigenicity and Aggressiveness of Ovarian Cancer Cells
Yu, S., Murph, M. M., Lu, Y., Liu, S., Hall, H. S., Liu, J., Stephens, C., Fang, X., & Mills, G. B. (2008) · JNCI: Journal of the National Cancer Institute · doi:10.1093/jnci/djn378 · Cited by 219
SIRENE: supervised inference of regulatory networks
Mordelet, F., & Vert, J.-P. (2008) · Bioinformatics · doi:10.1093/bioinformatics/btn273 · Cited by 191
From syndrome families to functional genomics
Brunner, H. G., & van Driel, M. A. (2004) · Nature Reviews Genetics · doi:10.1038/nrg1383 · Cited by 175 · cited 1×
“However, such kind of classifiers is actually built from a noisy negative set N as there can be unknown disease genes in”
BackgroundMordelet, F., & Vert, J.-P. (2011) · BMC Bioinformatics · doi:10.1186/1471-2105-12-389 · Cited by 175 · cited 1×
“However, such kind of classifiers is actually built from a noisy negative set N as there can be unknown disease genes in”
BackgroundAn integrated approach to inferring gene–disease associations in humans
Radivojac, P., Peng, K., Clark, W. T., Peters, B. J., Mohan, A., Boyle, S. M., & Mooney, S. D. (2008) · Proteins: Structure, Function, and Bioinformatics · doi:10.1002/prot.21989 · Cited by 175
Further understanding human disease genes by comparing with housekeeping genes and other genes
Tu, Z., Wang, L., Xu, M., Zhou, X. J., Chen, T., & Sun, F. (2006) · BMC Genomics · doi:10.1186/1471-2164-7-31 · Cited by 150
Afonyushkin, T., Oskolkova, O. V., Philippova, M., Resink, T. J., Erne, P., Binder, B. R., & Bochkov, V. N. (2010) · Arteriosclerosis, Thrombosis, and Vascular Biology · doi:10.1161/atvbaha.110.204354 · Cited by 143 · cited 4×
“Recent studies have revealed that genes associated with similar disorders have been shown to demonstrate higher probabilities of similar gene expression profiling (Ala et al., 2008), high functional similarities (Ideker and Sharan, 2008) and physical interactions between their gene products (Brunner et al., 2004; Goh…”
BackgroundInfluentialPrediction of Human Disease Genes by Human-Mouse Conserved Coexpression Analysis
Ala, U., Piro, R. M., Grassi, E., Damasco, C., Silengo, L., Oti, M., Provero, P., & Di Cunto, F. (2008) · PLoS Computational Biology · doi:10.1371/journal.pcbi.1000043 · Cited by 134
Learning gene regulatory networks from only positive and unlabeled data
Cerulo, L., Elkan, C., & Ceccarelli, M. (2010) · BMC Bioinformatics · doi:10.1186/1471-2105-11-228 · Cited by 131
LPA2 receptor mediates mitogenic signals in human colon cancer cells
Yun, C. C., Sun, H., Wang, D., Rusovici, R., Castleberry, A. M., Hall, R. A., & Shim, H. (2005) · American Journal of Physiology-Cell Physiology · doi:10.1152/ajpcell.00610.2004 · Cited by 129
Positive Unlabeled Learning for Data Stream Classification
Li, X.-L., Yu, P. S., Liu, B., & Ng, S.-K. (2009) · Proceedings of the 2009 SIAM International Conference on Data Mining · doi:10.1137/1.9781611972795.23 · Cited by 115
Self-citationInferring Gene-Phenotype Associations via Global Protein Complex Network Propagation
Yang, P., Li, X., Wu, M., Kwoh, C.-K., & Ng, S.-K. (2011) · PLoS ONE · doi:10.1371/journal.pone.0021502 · Cited by 109 · cited 1×
“However, such kind of classifiers is actually built from a noisy negative set N as there can be unknown disease genes in”
BackgroundSelf-citationTusell, S. M., Schittone, S. A., & Holmes, K. V. (2007) · Journal of Virology · doi:10.1128/jvi.01510-06 · Cited by 109
Rakeman, A. S., & Anderson, K. V. (2006) · Development · doi:10.1242/dev.02473 · Cited by 107
RBP2‐H1/JARID1B is a transcriptional regulator with a tumor suppressive potential in melanoma cells
Roesch, A., Mueller, A. M., Stempfl, T., Moehle, C., Landthaler, M., & VOGT, T. M. (2007) · International Journal of Cancer · doi:10.1002/ijc.23211 · Cited by 79
Respiratory chain complex I deficiency due to NDUFA12 mutations as a new cause of Leigh syndrome
Østergaard, E., Rodenburg, R. J., van den Brand, M., Thomsen, L. L., Dunø, M., Batbayli, M., Wibrand, F., & Nijtmans, L. G. J. (2011) · Journal of Medical Genetics · doi:10.1136/jmg.2011.088856 · Cited by 77
Maia, C. J., Socorro, S., Schmitt, F. C., & Santos, C. R. A. (2008) · Endocrine · doi:10.1007/s12020-008-9113-7 · Cited by 58
Takahashi, M., Takahashi, Y., Takahashi, K., Zolotaryov, F. N., Hong, K. S., Iida, K., Okimura, Y., Kaji, H., & Chihara, K. (2007) · Biochemical and Biophysical Research Communications · doi:10.1016/j.bbrc.2007.10.120 · Cited by 41
The Structures of the Human Calcium Channel α1 Subunit (CACNL1A2) and β Subunit (CACNLB3) Genes
Yamada, Y., Masuda, K., Li, Q., Ihara, Y., Kubota, A., Miura, T., Nakamura, K., Fujii, Y., Seino, S., & Seino, Y. (1995) · Genomics · doi:10.1006/geno.1995.1048 · Cited by 40
Learning to identify unexpected instances in the test set
Li, X., Liu, B., & Ng, S. (2007) · Cited by 35
Self-citationSmalter, A., Lei, S. F., & Chen, X. (2007) · IEEE International Conference on Bioinformatics and Biomedicine (BIBM) · doi:10.1109/bibm.2007.47 · Cited by 31
Visús, C., Ito, D., Dhir, R., Szczepański, M. J., Chang, Y. J., Latimer, J. J., Grant, S. G., & DeLeo, A. B. (2011) · Cancer Immunology Immunotherapy · doi:10.1007/s00262-011-1001-y · Cited by 23
Duax, W. L., Thomas, J. L., Pletnev, V. Z., Addlagatta, A., Huether, R., Habegger, L., & Weeks, C. M. (2005) · Annals of the New York Academy of Sciences · doi:10.1196/annals.1336.015 · Cited by 23
Matsuzaka, Y., Okamoto, K., Tsuji, H., Mabuchi, T., Ozawa, A., Tamiya, G., & Inoko, H. (2002) · Biochemical and Biophysical Research Communications · doi:10.1016/s0006-291x(02)02344-6 · Cited by 22
Mathematical Modeling of Phagocyte Chemotaxis toward and Adherence to Biomaterial Implants
Xue, J., Gao, J., & Tang, L. (2007) · IEEE International Conference on Bioinformatics and Biomedicine (BIBM) · doi:10.1109/bibm.2007.38 · Cited by 20
Stark, K., Törmä, H., & Oliw, E. H. (2006) · Prostaglandins & Other Lipid Mediators · doi:10.1016/j.prostaglandins.2005.12.003 · Cited by 20
Vajda, N. A., Brimacombe, K. R., LeMasters, K. E., & Ladd, A. N. (2009) · Developmental Dynamics · doi:10.1002/dvdy.22155 · Cited by 18
Cited by
Learning from positive and unlabeled data: a survey
Bekker, J., & Davis, J. (2020) · Machine Learning · doi:10.1007/s10994-020-05877-5 · Cited by 651
“Several PU methods were developed to this end: ProDiGe is a method based on bagging SVMs (Mordelet and Vert 2011, 2014), PUDI is also a weighted SVM method, but they have different weights for four identified groups of unlabeled examples: reliable negative, likely positive, likely negative and weakly negative (Yang et…”
MethodEnhancing the prediction of disease-gene associations with multimodal deep learning
Luo, P., Li, Y., Tian, L.-P., & Wu, F.-X. (2019) · Bioinformatics · doi:10.1093/bioinformatics/btz155 · Cited by 67
Binder, J., Ursu, O., Bologa, C., Jiang, S., Maphis, N., Dadras, S., Chisholm, D., Weick, J., Myers, O., Kumar, P., et al. (2022) · Communications Biology · doi:10.1038/s42003-022-03068-7 · Cited by 49
Formatting biological big data for modern machine learning in drug discovery
Duran‐Frigola, M., Fernández‐Torras, A., Bertoni, M., & Aloy, P. (2019) · WIREs Computational Molecular Science · doi:10.1002/wcms.1408 · Cited by 25
Metadata
- Authors
- Yang, P., Li, X.-L., Mei, J.-P., Kwoh, C.-K., & Ng, S.-K.
- Venue
- Bioinformatics
- Publisher
- Oxford University Press (OUP)
- Published
- 2012-08-24
- Volume
- 28
- Issue
- 20
- Pages
- 2640-2647
- ISSN
- 1367-4803 (ISSN-L)
- 1367-4811
- OpenAlex
- W2170283814
- Cited by
- 210
- References
- 59
- FWCI
- 2.09
- Type
- journal article
- Language
- en
- Open access
- hybrid
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.