Multitask Protein Function Prediction Through Task Dissimilarity
This work addresses protein function prediction for bioinformatics, but it is incremental as it modifies an existing multitask approach with a focus on dissimilarity.
The authors tackled the problem of automated protein function prediction, which is challenging due to hierarchical organization and data scarcity, by proposing a multitask learning algorithm that uses task dissimilarity instead of similarity to improve performance on unbalanced datasets, showing more stable results in experimental comparisons across three model organisms.
Automated protein function prediction is a challenging problem with distinctive features, such as the hierarchical organization of protein functions and the scarcity of annotated proteins for most biological functions. We propose a multitask learning algorithm addressing both issues. Unlike standard multitask algorithms, which use task (protein functions) similarity information as a bias to speed up learning, we show that dissimilarity information enforces separation of rare class labels from frequent class labels, and for this reason is better suited for solving unbalanced protein function prediction problems. We support our claim by showing that a multitask extension of the label propagation algorithm empirically works best when the task relatedness information is represented using a dissimilarity matrix as opposed to a similarity matrix. Moreover, the experimental comparison carried out on three model organism shows that our method has a more stable performance in both "protein-centric" and "function-centric" evaluation settings.