A feature selection strategy for gene expression time series experiments with hidden Markov models

Abstract Studies conducted in time series could be far more informative than those questioning at a specific moment in time. However, when it comes to genomic data, time points are sparse creating the need for a constant search for methods capable of extracting information out of experiments of this kind. We propose a feature selection algorithm embedded in a hidden Markov model applied to gene expression time course data on either single or even multiple biological conditions. For the latter, in a simple case-control study features or genes are selected under the assumption of no change over time for the control samples, while the case group must have at least one change. The proposed model reduces the feature space according to a two-state hidden Markov model. The two states define change/no-change in gene expression. Features are ranked in consonance with three scores: number of changes across time, magnitude of such changes and quality of replicates as a measure of how much they deviate from the mean. An important highlight is that this strategy overcomes the few samples limitation, common in genomic experiments through a process of data transformation and rearrangement. To prove this method, our strategy was applied to three publicly available data sets. Results show that feature domain is reduced to up to 90% leaving only few but relevant features yet with findings consistent to those previously reported. Moreover, our strategy proved to be robust, stable and working on studies where sample size is an issue otherwise. Hence, even with two biological replicates and/or three time points our method proves to work well. Author summaryThe variety of methods for the analysis of gene expression with longitudinal measurements proposed aim to classify genes according to their expression profile over time. Approaches focus on searching for genes with: large changes at any given time, multiple changes across the time series, similar/opposed profile or a focal time point where most changes occur among others. What differs between these methods is the mathematical way they reach those goals. In all cases, we deal with noisy data, small number of time points, limited replication and questionable dependencies. In this work, we present a review of the state-of-the-art of such methods and propose a novel algorithm that integrates a probabilistic graphical approach using Hidden Markov Models (HMMs), ideal for modeling sequential data convoluted with feature selection (FS) for gene classification. Its novel contribution or major innovation resides in how the method handles limited data points and low replication. This innovation is relevant because most genome-based studies face this same challenge, particularly longitudinal designs for which, sample size is a combination of time measurements and biological replicates.