Guylaine Poisson

Dr. Guylaine Poisson


Research Focus & Metrics

Academic Career Length

20Years (2007-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
1 June 2007 in Genomics, Proteomics & Bioinformatics
Verified
Journal Article
Open access
GPGuylaine Poisson
Guylaine Poisson
CCCedric Chauve
Cedric Chauve
XCXin Chen
Xin Chen
ABAnne Bergeron
Anne Bergeron
A glycosylphosphatidylinositol (GPI) anchor is a common but complex C-terminal post-translational modification of extracellular proteins in eukaryotes. Here we investigate the problem of correctly annotating GPI-anchored proteins for the growing number of sequences in public databases. We developed a computational system, called FragAnchor, based on the tandem use of a neural network (NN) and a hidden Markov model (HMM). Firstly, NN selects potential GPI-anchored proteins in a dataset, then HMM parses these potential GPI signals and refines the prediction by qualitative scoring. FragAnchor correctly predicted 91% of all the GPI-anchored proteins annotated in the Swiss-Prot database. In a large-scale analysis of 29 eukaryote proteomes, FragAnchor predicted that the percentage of highly probable GPI-anchored proteins is between 0.21% and 2.01%. The distinctive feature of FragAnchor, compared with other systems, is that it targets only the C-terminus of a protein, making it less sensitive to the background noise found in databases and possible incomplete protein sequences. Moreover, FragAnchor can be used to predict GPI-anchored proteins in all eukaryotes. Finally, by using qualitative scoring, the predictions combine both sensitivity and information content. The predictor is publicly available at http://navet.ics.hawaii.edu/~fraganchor/NNHMM/NNHMM.html.

5 October 2011 in BMC Bioinformatics
Verified
Journal Article
Open access
PDF file
MBMahdi Belcaid
Mahdi Belcaid
ABAnne Bergeron
Anne Bergeron
GPGuylaine Poisson
Guylaine Poisson
Background: A large family of viruses that infect bacteria, calledphages, is characterized by long tails used to inject DNA into their victims' cells. Thetape measure proteingot its name because the length of the corresponding gene is proportional to the length of the phage's tail: a fact shown by actually copying or splicing out parts of DNA in exemplar species. A natural question is whether there existunitsfor these tape measures, and if different tape measures have different units and lengths. Such units would allow us to retrace the evolution of tape measure proteins using their duplication/loss history. The vast number of sequenced phages genomes allows us to attack this problem with a comparative genomics approach.Results: Here we describe a subset of phages whose tape measure proteins contain variable numbers of an 11 amino acids sequence repeat, aligned with sequence similarity, structural properties, and simple arithmetics. This subset provides a unique opportunity for the combinatorial study of phage evolution, without the added uncertainties of multiple alignments, which are trivial in this case, or of protein functions, that are well established. We give a heuristic that reconstructs the duplication history of these sequences, using divergent strains to discriminate between mutations that occurred before and after speciation, or lineage divergence. The heuristic is based on an efficient algorithm that gives an exhaustive enumeration of all possible parsimonious reconstructions of the duplication/speciation history of a single nucleotide. Finally, we present a method that allows, when possible, to discriminate between duplication and loss events.Conclusions: Establishing the evolutionary history of viruses is difficult, in part due to extensive recombinations and gene transfers, and high mutation rates that often erase detectable similarity between homologous genes. In this paper, we introduce new tools to address this problem.

1 September 2010 in Journal of Computational Biology
Verified
Journal Article
Closed access
MBMahdi Belcaid
Mahdi Belcaid
ABAnne Bergeron
Anne Bergeron
GPGuylaine Poisson
Guylaine Poisson
Comparing the genomes of two closely related viruses often produces mosaics where nearly identical sequences alternate with sequences that are unique to each genome. When several closely related genomes are compared, the unique sequences are likely to be shared with third genomes, leading to virus mosaic communities. Here we present comparative analysis of sets of Staphylococcus aureus phages that share large identical sequences with up to three other genomes, and with different partners along their genomes. We introduce mosaic graphs to represent these complex recombination events, and use them to illustrate the breath and depth of sequence sharing: some genomes are almost completely made up of shared sequences, while genomes that share very large identical sequences can adopt alternate functional modules. Mosaic graphs also allow us to identify breakpoints that could eventually be used for the construction of recombination networks. These findings have several implications on phage metagenomics assembly, on the horizontal gene transfer paradigm, and more generally on the understanding of the composition and evolutionary dynamics of virus communities.

Verified
Journal Article
Open access
PDF file
MMMark S. Menor
Mark S. Menor
KBKyungim Baek
Kyungim Baek
GPGuylaine Poisson
Guylaine Poisson
The discovery of novel microRNA (miRNA) and piwi-interacting RNA (piRNA) is an important task for the understanding of many biological processes. Most of the available miRNA and piRNA identification methods are dependent on the availability of the organism’s genome sequence and the quality of its annotation. Therefore, an efficient prediction method based solely on the short RNA reads and requiring no genomic information is highly desirable. In this study, we propose an approach that relies primarily on the nucleotide composition of the read and does not require reference genomes of related species for prediction. Using an empirical Bayesian kernel method and the error correcting output codes framework, compact models suitable for large-scale analyses are built on databases of known mature miRNAs and piRNAs. We found that the usage of an L1-based Gaussian kernel can double the true positive rate compared to the standard L2-based Gaussian kernel. Our approach can increase the true positive rate by at most 60% compared to the existing piRNA predictor based on the analysis of a hold-out test set. Using experimental data, we also show that our approach can detect about an order of magnitude or more known miRNAs than the mature miRNA predictor, miRPlex.

15 February 2013 in BMC Genomics
Verified
Journal Article
Open access
PDF file
MMMark Menor
Mark Menor
KBKyungim Baek
Kyungim Baek
GPGuylaine Poisson
Guylaine Poisson
Background: Classification is the problem of assigning each input object to one of a finite number of classes. This problem has been extensively studied in machine learning and statistics, and there are numerous applications to bioinformatics as well as many other fields. Building a multiclass classifier has been a challenge, where the direct approach of altering the binary classification algorithm to accommodate more than two classes can be computationally too expensive. Hence the indirect approach of using binary decomposition has been commonly used, in which retrieving the class posterior probabilities from the set of binary posterior probabilities given by the individual binary classifiers has been a major issue. Methods: In this work, we present an extension of a recently introduced probabilistic kernel-based learning algorithm called the Classification Relevance Units Machine (CRUM) to the multiclass setting to increase its applicability. The extension is achieved under the error correcting output codes framework. The probabilistic outputs of the binary CRUMs are preserved using a proposed linear-time decoding algorithm, an alternative to the generalized Bradley-Terry (GBT) algorithm whose application to large-scale prediction settings is prohibited by its computational complexity. The resulting classifier is called the Multiclass Relevance Units Machine (McRUM). Results: The evaluation of McRUM on a variety of real small-scale benchmark datasets shows that our proposed Naïve decoding algorithm is computationally more efficient than the GBT algorithm while maintaining a similar level of predictive accuracy. Then a set of experiments on a larger scale dataset for small ncRNA classification have been conducted with Naïve McRUM and compared with the Gaussian and linear SVM. Although McRUM's predictive performance is slightly lower than the Gaussian SVM, the results show that the similar level of true positive rate can be achieved by sacrificing false positive rate slightly. Furthermore, McRUM is computationally more efficient than the SVM, which is an important factor for large-scale analysis. Conclusions: We have proposed McRUM, a multiclass extension of binary CRUM. McRUM with Naïve decoding algorithm is computationally efficient in run-time and its predictive performance is comparable to the well-known SVM, showing its potential in solving large-scale multiclass problems in bioinformatics and other fields of study.

1 December 2012 in ACM SIGAPP Applied Computing Review
Verified
Journal Article
Open access
MMMark Menor
Mark Menor
KBKyungim Baek
Kyungim Baek
GPGuylaine Poisson
Guylaine Poisson
Phosphorylation is an important post-translational modification of proteins that is essential to the regulation of many cellular processes. Although most of the phosphorylation sites discovered in protein sequences have been identified experimentally, the in vivo and in vitro discovery of the sites is an expensive, time-consuming and laborious task. Therefore, the development of computational methods for prediction of protein phosphorylation sites has drawn considerable attention. In this work, we present a kernel-based probabilistic Classification Relevance Units Machine (CRUM) for in silico phosphorylation site prediction. In comparison with the popular Support Vector Machine (SVM) CRUM shows comparable predictive performance and yet provides a more parsimonious model. This is desirable since it leads to a reduction in prediction run-time, which is important in predictions on large-scale data. Furthermore, the CRUM training algorithm has lower run-time and memory complexity and has a simpler parameter selection scheme than the Relevance Vector Machine (RVM) learning algorithm. To further investigate the viability of using CRUM in phosphorylation site prediction, we construct multiple CRUM predictors using different combinations of three phosphorylation site features -- BLOSUM encoding, disorder, and amino acid composition. The predictors are evaluated through cross-validation and the results show that CRUM with BLOSUM feature is among the best performing CRUM predictors in both cross-validation and benchmark experiments. A comparative study with existing prediction tools in an independent benchmark experiment suggests possible direction for further improving the predictive performance of CRUM predictors.