A hierarchical multi-label classification ant colony algorithm for protein function prediction

Abstract

This paper proposes a novel Ant Colony Optimisation algorithm (ACO) tailored for the hierarchical multilabel classification problem of protein function prediction. This problem is a very active research field, given the large increase in the number of uncharacterised proteins available for analysis and the importance of determining their functions in order to improve the current biological knowledge. Since it is known that a protein can perform more than one function and many protein functional-definition schemes are organised in a hierarchical structure, the classification problem in this case is an instance of a hierarchical multi-label problem. In this type of problem, each example may belong to multiple class labels and class labels are organised in a hierarchical structure-either a tree or a directed acyclic graph (DAG) structure. It presents a more complex problem than conventional flat classification, given that the classification algorithm has to take into account hierarchical relationships between class labels and be able to predict multiple class labels for the same example. The proposed ACO algorithm discovers an ordered list of hierarchical multi-label classification rules. It is evaluated on sixteen challenging bioinformatics data sets involving hundreds or thousands of class labels to be predicted and compared against state-of-theart decision tree induction algorithms for hierarchical multilabel classification.

Probabilistic Clustering for Hierarchical Multi-Label Classification of Protein Functions

“…The authors later extended this method [23] to allow multi-label classification, considering both tree-and DAG-structured hierarchies.…”

“…The three methods are detailed in [31]. We also employ the Ant Colony Optimization-based method hmAnt-Miner [23], which is a global method that obtained competitive results when compared to Clus-HMC. We decided to select these algorithms because they were all applied to the same protein function prediction datasets used in the experiments (both for tree and DAG structures).…”

A Hierarchical Multi-Label Classification Algorithm for Gene Function Prediction

“…Firstly, as the hierarchy is traversed down to the leaves, classes at lower levels of the hierarchy usually have very fewer positive instances, requiring base classifiers at these hierarchy levels to be learned from very skewed data sets [19]. In such imbalanced data set learning scenarios, standard classifiers might be overwhelmed by the majority class and can not detect the minority one.…”

Predicting protein function via multi-label supervised topic model on gene ontology

“…Meanwhile, from the view of protein function prediction, a protein annotated with several GO terms can be viewed as a sample and each GO term corresponds to a label. Thus, the solutions of multi-label classification have great potential to protein function prediction [10][11][12][13][14]: Celine Vens et al [15] discussed three hierarchical multi-label classifiers, which are based on the decision tree algorithm; then, the experiments were carried on 24 datasets from functional genomics; Yu [16] proposed a multiple kernels (ProMK) method to process multiple heterogeneous protein data sources for predicting protein functions; a novel ant colony algorithm was proposed in [17] for hierarchical multi-label classification. Although good results of multi-label classification are achieved by methods mentioned above, the outputs of them are the corresponding label sets for samples in testing set, and the relationships between individual attributes and their most appropriate labels are unknown.…”