Abstract
Motivation: Simple bioinformatic tools are frequently used to analyse time-series datasets regardless of their ability to deal with transient phenomena, limiting the meaningful information that may be extracted from them. This situation requires the development and exploitation of tailor-made, easy-to-use and flexible tools designed specifically for the analysis of time-series datasets.Results: We present a novel statistical application called CLUSTERnGO, which uses a model-based clustering algorithm that fulfils this need. This algorithm involves two components of operation. Component 1 constructs a Bayesian non-parametric model (Infinite Mixture of Piecewise Linear Sequences) and Component 2, which applies a novel clustering methodology (Two-Stage Clustering). The software can also assign biological meaning to the identified clusters using an appropriate ontology. It applies multiple hypothesis testing to report the significance of these enrichments. The algorithm has a four-phase pipeline. The application can be executed using either command-line tools or a user-friendly Graphical User Interface. The latter has been developed to address the needs of both specialist and non-specialist users. We use three diverse test cases to demonstrate the flexibility of the proposed strategy. In all cases, CLUSTERnGO not only outperformed existing algorithms in assigning unique GO term enrichments to the identified clusters, but also revealed novel insights regarding the biological systems examined, which were not uncovered in the original publications.Availability and implementation: The C++ and QT source codes, the GUI applications for Windows, OS X and Linux operating systems and user manual are freely available for download under the GNU GPL v3 license at http://www.cmpe.boun.edu.tr/content/CnG.Contact: sgo24@cam.ac.ukSupplementary information: Supplementary data are available at Bioinformatics online.
Highlights
High-throughput technologies in the life sciences generate massive amounts of information by allowing the measurement of thousands of entities simultaneously
The performance of CnG was evaluated by comparing the extent of biological insight gained employing this methodology to that gained by two predecessor model-based algorithms, Chinese Restaurant Cluster (CRC) and Gaussian infinite mixture model (GIMM)
We carried out an internal evaluation of the clustering results to assess the quality of the set of clusters obtained from CnG in comparison to CRC and GIMM
Summary
High-throughput technologies in the life sciences generate massive amounts of information by allowing the measurement of thousands of entities simultaneously. A two-stage complete linkage clustering procedure was employed to identify the patterns in the data Unlike its predecessors, this simple and effective approach can address all of the following issues simultaneously: (i) it allows the user to construct their own model, which would integratively take into account both the design of the experiment and the collected data, prior to analysis, (ii) it has a deterministic clustering output, despite its probabilistic approach introduced by two-stage clustering, (iii) it takes into account the differences and the similarities in both the profiles and the magnitudes of expression, (iv) it is suitable for or unequally sampled long or short time-series datasets, (v) it does not require an a priori knowledge or assumption on the number of clusters that will be identified at the end of the process, (vi) it allows the assignment of the same gene into different clusters, i.e. overlapping clusters, minimizing the loss of biological information hidden in the dataset introduced by two-stage clustering, and (vii) it has a very friendly GUI suitable for both specialist and non-specialist users despite the rigorous computational procedures running in the background. We test the applicability of our approach on three independent published biological datasets, which are different in size, the level of gene expression under investigation, the temporal experimental design, the presence of replicates, as well as the level of complexity of the model organism and demonstrate that our algorithm brings substantial novel insight into the systems under investigation, which was previously not reported and outperforms its predecessors in doing so
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
