Predicting gene function using hierarchical multi-label decision tree ensembles

Abstract

BackgroundS. cerevisiae, A. thaliana and M. musculus are well-studied organisms in biology and the sequencing of their genomes was completed many years ago. It is still a challenge, however, to develop methods that assign biological functions to the ORFs in these genomes automatically. Different machine learning methods have been proposed to this end, but it remains unclear which method is to be preferred in terms of predictive performance, efficiency and usability.ResultsWe study the use of decision tree based models for predicting the multiple functions of ORFs. First, we describe an algorithm for learning hierarchical multi-label decision trees. These can simultaneously predict all the functions of an ORF, while respecting a given hierarchy of gene functions (such as FunCat or GO). We present new results obtained with this algorithm, showing that the trees found by it exhibit clearly better predictive performance than the trees found by previously described methods. Nevertheless, the predictive performance of individual trees is lower than that of some recently proposed statistical learning methods. We show that ensembles of such trees are more accurate than single trees and are competitive with state-of-the-art statistical learning and functional linkage methods. Moreover, the ensemble method is computationally efficient and easy to use.ConclusionsOur results suggest that decision tree based methods are a state-of-the-art, efficient and easy-to-use approach to ORF function prediction.

Extensive complementarity between gene function prediction methods

“…CLUS-HMC can exploit the hierarchical relationships in GO to achieve higher predictive performance (Blockeel et al, 2006;Vens et al, 2008) and was previously used for AFP tasks (Schietgat et al, 2010;Slavkov et al, 2010;Škunca et al, 2013). For each OG and GO term pair, the classifier outputs a score ranging between 0 and 1 that indicates confidence in assignment of that function to the OG.…”

Fast Label Embeddings via Randomized Linear Algebra

“…when composed with online learning [11]; Bayesian learning [10]; support vector machines [6]; and decision tree ensembles [39]. Our embedding approach neither requires nor exploits such side information, and is therefore applicable to different scenarios, but is potentially suboptimal when side information is present.…”

ACO-Based Bayesian Network Ensembles for the Hierarchical Classification of Ageing-Related Proteins

“…Table 1 shows how our work fits into the context of related work on hierarchical classification according to these two issues. [15] In addition to the use of our ABC-Miner for hierarchical protein function prediction in a new ageing-related dataset, what is novel in this work are the various methods we employ to combine the outputs of an ensemble's classifiers built with different protein representations in hierarchical classification. As shown in Table 1, three of the four cells marked with the keyword "this work" involve a new combination of the technique of building classifiers with different protein representations (proposed in [18]) with a technique for combining the classifiers' predictions.…”