Sven Degroeve

Large-scale structural analysis of the core promoter in mammalian and plant genomes

DNA encodes at least two independent levels of functional information. The first level is for encoding proteins and sequence targets for DNA-binding factors, while the second one is contained in the physical and structural properties of the DNA molecule itself. Although the physical and structural properties are ultimately determined by the nucleotide sequence itself, the cell exploits these properties in a way in which the sequence itself plays no role other than to support or facilitate certain spatial structures. In this work, we focus on these structural properties, comparing them between different organisms and assessing their ability to describe the core promoter. We prove the existence of distinct types of core promoters, based on a clustering of their structural profiles. These results indicate that the structural profiles are much conserved within plants (Arabidopsis and rice) and animals (human and mouse), but differ considerably between plants and animals. Furthermore, we demonstrate that these structural profiles can be an alternative way of describing the core promoter, in addition to more classical motif or IUPAC-based approaches. Using the structural profiles as discriminatory elements to separate promoter regions from non-promoter regions, reliable models can be built to identify core-promoter regions using a strictly computational approach.

SpliceMachine: predicting splice sites from high-dimensional local context representations

MOTIVATION In this age of complete genome sequencing, finding the location and structure of genes is crucial for further molecular research. The accurate prediction of intron boundaries largely facilitates the correct prediction of gene structure in nuclear genomes. Many tools for localizing these boundaries on DNA sequences have been developed and are available to researchers through the internet. Nevertheless, these tools still make many false positive predictions. RESULTS This manuscript presents a novel publicly available splice site prediction tool named SpliceMachine that (i) shows state-of-the-art prediction performance on Arabidopsis thaliana and human sequences, (ii) performs a computationally fast annotation and (iii) can be trained by the user on its own data. AVAILABILITY Results, figures and software are available at http://www.bioinformatics.psb.ugent.be/supplementary_data/ CONTACT sven.degroeve@psb.ugent.be; yves.vandepeer@psb.ugent.be.

Feature selection for splice site prediction: A new method using EDA-based feature ranking

BackgroundThe identification of relevant biological features in large and complex datasets is an important step towards gaining insight in the processes underlying the data. Other advantages of feature selection include the ability of the classification system to attain good or even better solutions using a restricted subset of features, and a faster classification. Thus, robust methods for fast feature selection are of key importance in extracting knowledge from complex biological data.ResultsIn this paper we present a novel method for feature subset selection applied to splice site prediction, based on estimation of distribution algorithms, a more general framework of genetic algorithms. From the estimated distribution of the algorithm, a feature ranking is derived. Afterwards this ranking is used to iteratively discard features. We apply this technique to the problem of splice site prediction, and show how it can be used to gain insight into the underlying biological process of splicing.ConclusionWe show that this technique proves to be more robust than the traditional use of estimation of distribution algorithms for feature selection: instead of returning a single best subset of features (as they normally do) this method provides a dynamical view of the feature selection process, like the traditional sequential wrapper methods. However, the method is faster than the traditional techniques, and scales better to datasets described by a large number of features.

Analysis of the resolution limitations of peptide identification algorithms.

Proteome identification using peptide-centric proteomics techniques is a routinely used analysis technique. One of the most powerful and popular methods for the identification of peptides from MS/MS spectra is protein database matching using search engines. Significance thresholding through false discovery rate (FDR) estimation by target/decoy searches is used to ensure the retention of predominantly confident assignments of MS/MS spectra to peptides. However, shortcomings have become apparent when such decoy searches are used to estimate the FDR. To study these shortcomings, we here introduce a novel kind of decoy database that contains isobaric mutated versions of the peptides that were identified in the original search. Because of the supervised way in which the entrapment sequences are generated, we call this a directed decoy database. Since the peptides found in our directed decoy database are thus specifically designed to look quite similar to the forward identifications, the limitations of the existing search algorithms in making correct calls in such strongly confusing situations can be analyzed. Interestingly, for the vast majority of confidently identified peptide identifications, a directed decoy peptide-to-spectrum match can be found that has a better or equal match score than the forward match score, highlighting an important issue in the interpretation of peptide identifications in present-day high-throughput proteomics.

A posteriori quality control for the curation and reuse of public proteomics data

Proteomics is a rapidly expanding field encompassing a multitude of complex techniques and data types. To date much effort has been devoted to achieving the highest possible coverage of proteomes with the aim to inform future developments in basic biology as well as in clinical settings. As a result, growing amounts of data have been deposited in publicly available proteomics databases. These data are in turn increasingly reused for orthogonal downstream purposes such as data mining and machine learning. These downstream uses however, need ways to a posteriori validate whether a particular data set is suitable for the envisioned purpose. Furthermore, the (semi‐)automatic curation of repository data is dependent on analyses that can highlight misannotation and edge conditions for data sets. Such curation is an important prerequisite for efficient proteomics data reuse in the life sciences in general. We therefore present here a selection of quality control metrics and approaches for the a posteriori detection of potential issues encountered in typical proteomics data sets. We illustrate our metrics by relying on publicly available data from the Proteomics Identifications Database (PRIDE), and simultaneously show the usefulness of the large body of PRIDE data as a means to derive empirical background distributions for relevant metrics.

MS2PIP: a tool for MS/MS peak intensity prediction.

MOTIVATION Tandem mass spectrometry provides the means to match mass spectrometry signal observations with the chemical entities that generated them. The technology produces signal spectra that contain information about the chemical dissociation pattern of a peptide that was forced to fragment using methods like collision-induced dissociation. The ability to predict these MS(2) signals and to understand this fragmentation process is important for sensitive high-throughput proteomics research. RESULTS We present a new tool called MS(2)PIP for predicting the intensity of the most important fragment ion signal peaks from a peptide sequence. MS(2)PIP pre-processes a large dataset with confident peptide-to-spectrum matches to facilitate data-driven model induction using a random forest regression learning algorithm. The intensity predictions of MS(2)PIP were evaluated on several independent evaluation sets and found to correlate significantly better with the observed fragment-ion intensities as compared with the current state-of-the-art PeptideART tool. AVAILABILITY MS(2)PIP code is available for both training and predicting at http://compomics.com/.

Combining quantitative proteomics data processing workflows for greater sensitivity

We here describe a normalization method to combine quantitative proteomics data. By merging the output of two popular quantification software packages, we obtained a 20% increase (on average) in the number of quantified human proteins without suffering from a loss of quality. Our integrative workflow is freely available through our user-friendly, open-source Rover software (http://compomics-rover.googlecode.com/).

Predicting tryptic cleavage from proteomics data using decision tree ensembles

Trypsin is the workhorse protease in mass spectrometry-based proteomics experiments and is used to digest proteins into more readily analyzable peptides. To identify these peptides after mass spectrometric analysis, the actual digestion has to be mimicked as faithfully as possible in silico. In this paper we introduce CP-DT (Cleavage Prediction with Decision Trees), an algorithm based on a decision tree ensemble that was learned on publicly available peptide identification data from the PRIDE repository. We demonstrate that CP-DT is able to accurately predict tryptic cleavage: tests on three independent data sets show that CP-DT significantly outperforms the Keil rules that are currently used to predict tryptic cleavage. Moreover, the trees generated by CP-DT can make predictions efficiently and are interpretable by domain experts.

Towards a human proteomics atlas

Proteomics research has taken up an increasingly important role in life sciences over the past few years. Due to a strong push from publishers and funders alike, the community has also started to freely share its data in earnest, making use of public repositories such as the highly popular PRIDE database at EMBL-EBI. Reuse of these publicly available data has so far been confined to rather specific, targeted reanalyses, but this limited reuse is set to expand dramatically as repositories continue to grow exponentially. Examples of large-scale reuse are readily found in other omics disciplines, where more comprehensive public data have already accumulated over longer periods. Here, a typical example of integrative data reuse is provided by the construction of so-called expression atlases. We here therefore investigate the issues involved in using the human data currently stored in the PRIDE database to construct a robust, tissue-specific protein expression atlas from tandem-MS based label-free quantification.

Digging into acceptor splice site prediction: an iterative feature selection approach

Feature selection techniques are often used to reduce data dimensionality, increase classification performance, and gain insight into the processes that generated the data. In this paper, we describe an iterative procedure of feature selection and feature construction steps, improving the classification of acceptor splice sites, an important subtask of gene prediction. We show that acceptor prediction can benefit from feature selection, and describe how feature selection techniques can be used to gain new insights in the classification of acceptor sites. This is illustrated by the identification of a new, biologically motivated feature: the AG-scanning feature.The results described in this paper contribute both to the domain of gene prediction, and to research in feature selection techniques, describing a new wrapper based feature weighting method that aids in knowledge discovery when dealing with complex datasets.