Limits...
Pathway activity inference for multiclass disease classification through a mathematical programming optimisation framework.

Yang L, Ainali C, Tsoka S, Papageorgiou LG - BMC Bioinformatics (2014)

Bottom Line: We show that not only does it improve classification accuracy, but it can also perform well in multiclass disease datasets, a limitation of other approaches from the literature.Desirable features of the model include the ability to control the maximum number of genes that may participate in determining pathway activity, which may be pre-specified by the user.Overall, this work highlights the potential of building pathway-based multi-phenotype classifiers for accurate disease diagnosis and prognosis problems.

View Article: PubMed Central - PubMed

Affiliation: Centre for Process Systems Engineering, Department of Chemical Engineering, University College London, London, WC1E 7JE, UK. lingjian.yang.10@ucl.ac.uk.

ABSTRACT

Background: Applying machine learning methods on microarray gene expression profiles for disease classification problems is a popular method to derive biomarkers, i.e. sets of genes that can predict disease state or outcome. Traditional approaches where expression of genes were treated independently suffer from low prediction accuracy and difficulty of biological interpretation. Current research efforts focus on integrating information on protein interactions through biochemical pathway datasets with expression profiles to propose pathway-based classifiers that can enhance disease diagnosis and prognosis. As most of the pathway activity inference methods in literature are either unsupervised or applied on two-class datasets, there is good scope to address such limitations by proposing novel methodologies.

Results: A supervised multiclass pathway activity inference method using optimisation techniques is reported. For each pathway expression dataset, patterns of its constituent genes are summarised into one composite feature, termed pathway activity, and a novel mathematical programming model is proposed to infer this feature as a weighted linear summation of expression of its constituent genes. Gene weights are determined by the optimisation model, in a way that the resulting pathway activity has the optimal discriminative power with regards to disease phenotypes. Classification is then performed on the resulting low-dimensional pathway activity profile.

Conclusions: The model was evaluated through a variety of published gene expression profiles that cover different types of disease. We show that not only does it improve classification accuracy, but it can also perform well in multiclass disease datasets, a limitation of other approaches from the literature. Desirable features of the model include the ability to control the maximum number of genes that may participate in determining pathway activity, which may be pre-specified by the user. Overall, this work highlights the potential of building pathway-based multi-phenotype classifiers for accurate disease diagnosis and prognosis problems.

Show MeSH

Related in: MedlinePlus

Sensitivity analysis of parameterNoGfor DIGS model with SMO (A) and NN (B) classifiers. For each of the 8 datasets, the proposed DIGS model is applied to infer pathway activity while setting NoG, i.e. the maximum number of member genes in a pathway allowed to have non-zero weights, to 5, 10, 15 and 20. In addition, DIGS model is also applied with NoG set to equal to the number of available member genes in a pathway, i.e. all member genes can take non-zero weights to construct pathway activity. A classifier is trained using the pathway activity profiles and tests the prediction accuracy. For both SMO (A) and NN (B) classifiers, it is clear that the proposed DIGS model is robust to the parameter NoG during the tested ranged 5 to 20. Furthermore, constraining the maximum number of active constituent genes appears to generally improve classification accuracy as DIGS_ALL usually leads to lower prediction rate compared with the others.
© Copyright Policy - open-access
Related In: Results  -  Collection

License 1 - License 2
getmorefigures.php?uid=PMC4269079&req=5

Fig2: Sensitivity analysis of parameterNoGfor DIGS model with SMO (A) and NN (B) classifiers. For each of the 8 datasets, the proposed DIGS model is applied to infer pathway activity while setting NoG, i.e. the maximum number of member genes in a pathway allowed to have non-zero weights, to 5, 10, 15 and 20. In addition, DIGS model is also applied with NoG set to equal to the number of available member genes in a pathway, i.e. all member genes can take non-zero weights to construct pathway activity. A classifier is trained using the pathway activity profiles and tests the prediction accuracy. For both SMO (A) and NN (B) classifiers, it is clear that the proposed DIGS model is robust to the parameter NoG during the tested ranged 5 to 20. Furthermore, constraining the maximum number of active constituent genes appears to generally improve classification accuracy as DIGS_ALL usually leads to lower prediction rate compared with the others.

Mentions: Here, the DIGS model is applied to infer pathway activity with NoG set to 5, 10, 15 and 20, followed by training and testing using a range of classifiers for each microarray dataset. As a comparison, DIGS is also run with NoG set equal to the number of member genes for each pathway, so as to allow all member genes in a pathway to take non-zero weights for pathway activity inference. The prediction rates achieved by these different values of NoG are denoted by DIGS_5, DIGS_10, DIGS_15, DIGS_20 and DIGS_ALL and are shown in FigureĀ 2A and B with SMO and NN classifiers and other classifiers in Additional file 2.Figure 2


Pathway activity inference for multiclass disease classification through a mathematical programming optimisation framework.

Yang L, Ainali C, Tsoka S, Papageorgiou LG - BMC Bioinformatics (2014)

Sensitivity analysis of parameterNoGfor DIGS model with SMO (A) and NN (B) classifiers. For each of the 8 datasets, the proposed DIGS model is applied to infer pathway activity while setting NoG, i.e. the maximum number of member genes in a pathway allowed to have non-zero weights, to 5, 10, 15 and 20. In addition, DIGS model is also applied with NoG set to equal to the number of available member genes in a pathway, i.e. all member genes can take non-zero weights to construct pathway activity. A classifier is trained using the pathway activity profiles and tests the prediction accuracy. For both SMO (A) and NN (B) classifiers, it is clear that the proposed DIGS model is robust to the parameter NoG during the tested ranged 5 to 20. Furthermore, constraining the maximum number of active constituent genes appears to generally improve classification accuracy as DIGS_ALL usually leads to lower prediction rate compared with the others.
© Copyright Policy - open-access
Related In: Results  -  Collection

License 1 - License 2
Show All Figures
getmorefigures.php?uid=PMC4269079&req=5

Fig2: Sensitivity analysis of parameterNoGfor DIGS model with SMO (A) and NN (B) classifiers. For each of the 8 datasets, the proposed DIGS model is applied to infer pathway activity while setting NoG, i.e. the maximum number of member genes in a pathway allowed to have non-zero weights, to 5, 10, 15 and 20. In addition, DIGS model is also applied with NoG set to equal to the number of available member genes in a pathway, i.e. all member genes can take non-zero weights to construct pathway activity. A classifier is trained using the pathway activity profiles and tests the prediction accuracy. For both SMO (A) and NN (B) classifiers, it is clear that the proposed DIGS model is robust to the parameter NoG during the tested ranged 5 to 20. Furthermore, constraining the maximum number of active constituent genes appears to generally improve classification accuracy as DIGS_ALL usually leads to lower prediction rate compared with the others.
Mentions: Here, the DIGS model is applied to infer pathway activity with NoG set to 5, 10, 15 and 20, followed by training and testing using a range of classifiers for each microarray dataset. As a comparison, DIGS is also run with NoG set equal to the number of member genes for each pathway, so as to allow all member genes in a pathway to take non-zero weights for pathway activity inference. The prediction rates achieved by these different values of NoG are denoted by DIGS_5, DIGS_10, DIGS_15, DIGS_20 and DIGS_ALL and are shown in FigureĀ 2A and B with SMO and NN classifiers and other classifiers in Additional file 2.Figure 2

Bottom Line: We show that not only does it improve classification accuracy, but it can also perform well in multiclass disease datasets, a limitation of other approaches from the literature.Desirable features of the model include the ability to control the maximum number of genes that may participate in determining pathway activity, which may be pre-specified by the user.Overall, this work highlights the potential of building pathway-based multi-phenotype classifiers for accurate disease diagnosis and prognosis problems.

View Article: PubMed Central - PubMed

Affiliation: Centre for Process Systems Engineering, Department of Chemical Engineering, University College London, London, WC1E 7JE, UK. lingjian.yang.10@ucl.ac.uk.

ABSTRACT

Background: Applying machine learning methods on microarray gene expression profiles for disease classification problems is a popular method to derive biomarkers, i.e. sets of genes that can predict disease state or outcome. Traditional approaches where expression of genes were treated independently suffer from low prediction accuracy and difficulty of biological interpretation. Current research efforts focus on integrating information on protein interactions through biochemical pathway datasets with expression profiles to propose pathway-based classifiers that can enhance disease diagnosis and prognosis. As most of the pathway activity inference methods in literature are either unsupervised or applied on two-class datasets, there is good scope to address such limitations by proposing novel methodologies.

Results: A supervised multiclass pathway activity inference method using optimisation techniques is reported. For each pathway expression dataset, patterns of its constituent genes are summarised into one composite feature, termed pathway activity, and a novel mathematical programming model is proposed to infer this feature as a weighted linear summation of expression of its constituent genes. Gene weights are determined by the optimisation model, in a way that the resulting pathway activity has the optimal discriminative power with regards to disease phenotypes. Classification is then performed on the resulting low-dimensional pathway activity profile.

Conclusions: The model was evaluated through a variety of published gene expression profiles that cover different types of disease. We show that not only does it improve classification accuracy, but it can also perform well in multiclass disease datasets, a limitation of other approaches from the literature. Desirable features of the model include the ability to control the maximum number of genes that may participate in determining pathway activity, which may be pre-specified by the user. Overall, this work highlights the potential of building pathway-based multi-phenotype classifiers for accurate disease diagnosis and prognosis problems.

Show MeSH
Related in: MedlinePlus