Background

Machine learning is a powerful approach for describing and predicting classes in microarray data. Although several comparative studies have investigated the relative performance of various machine learning methods, these often do not account for the fact that performance (e.g. error rate) is a result of a series of analysis steps of which the most important are data normalization, gene selection and machine learning.

Results

In this study, we used seven previously published cancer-related microarray data sets to compare the effects on classification performance of five normalization methods, three gene selection methods with 21 different numbers of selected genes and eight machine learning methods. Performance in term of error rate was rigorously estimated by repeatedly employing a double cross validation approach. Since performance varies greatly between data sets, we devised an analysis method that first compares methods within individual data sets and then visualizes the comparisons across data sets. We discovered both well performing individual methods and synergies between different methods.

Conclusion

Support Vector Machines with a radial basis kernel, linear kernel or polynomial kernel of degree 2 all performed consistently well across data sets. We show that there is a synergistic relationship between these methods and gene selection based on the T-test and the selection of a relatively high number of genes. Also, we find that these methods benefit significantly from using normalized data, although it is hard to draw general conclusions about the relative performance of different normalization procedures.

Background

Green plant leaves have always fascinated biologists as hosts for photosynthesis and providers of basic energy to many food webs. Today, comprehensive databases of gene expression data enable us to apply increasingly more advanced computational methods for reverse-engineering the regulatory network of leaves, and to begin to understand the gene interactions underlying complex emergent properties related to stress-response and development. These new systems biology methods are now also being applied to organisms such as Populus, a woody perennial tree, in order to understand the specific characteristics of these species.

Results

We present a systems biology model of the regulatory network of Populus leaves. The network is reverse-engineered from promoter information and expression profiles of leaf-specific genes measured over a large set of conditions related to stress and developmental. The network model incorporates interactions between regulators, such as synergistic and competitive relationships, by evaluating increasingly more complex regulatory mechanisms, and is therefore able to identify new regulators of leaf development not found by traditional genomics methods based on pair-wise expression similarity. The approach is shown to explain available gene function information and to provide robust prediction of expression levels in new data. We also use the predictive capability of the model to identify condition-specific regulation as well as conserved regulation between Populus and Arabidopsis.

Conclusions

We outline a computationally inferred model of the regulatory network of Populus leaves, and show how treating genes as interacting, rather than individual, entities identifies new regulators compared to traditional genomics analysis. Although systems biology models should be used with care considering the complexity of regulatory programs and the limitations of current genomics data, methods describing interactions can provide hypotheses about the underlying cause of emergent properties and are needed if we are to identify target genes other than those constituting the "low hanging fruit" of genomic analysis.

Background

Cluster analysis, and in particular hierarchical clustering, is widely used to extract information from gene expression data. The aim is to discover new classes, or sub-classes, of either individuals or genes. Performing a cluster analysis commonly involve decisions on how to; handle missing values, standardize the data and select genes. In addition, pre-processing, involving various types of filtration and normalization procedures, can have an effect on the ability to discover biologically relevant classes. Here we consider cluster analysis in a broad sense and perform a comprehensive evaluation that covers several aspects of cluster analyses, including normalization.

Result

We evaluated 2780 cluster analysis methods on seven publicly available 2-channel microarray data sets with common reference designs. Each cluster analysis method differed in data normalization (5 normalizations were considered), missing value imputation (2), standardization of data (2), gene selection (19) or clustering method (11). The cluster analyses are evaluated using known classes, such as cancer types, and the adjusted Rand index. The performances of the different analyses vary between the data sets and it is difficult to give general recommendations. However, normalization, gene selection and clustering method are all variables that have a significant impact on the performance. In particular, gene selection is important and it is generally necessary to include a relatively large number of genes in order to get good performance. Selecting genes with high standard deviation or using principal component analysis are shown to be the preferred gene selection methods. Hierarchical clustering using Ward's method, k-means clustering and Mclust are the clustering methods considered in this paper that achieves the highest adjusted Rand. Normalization can have a significant positive impact on the ability to cluster individuals, and there are indications that background correction is preferable, in particular if the gene selection is successful. However, this is an area that needs to be studied further in order to draw any general conclusions.

Conclusions

The choice of cluster analysis, and in particular gene selection, has a large impact on the ability to cluster individuals correctly based on expression profiles. Normalization has a positive effect, but the relative performance of different normalizations is an area that needs more research. In summary, although clustering, gene selection and normalization are considered standard methods in bioinformatics, our comprehensive analysis shows that selecting the right methods, and the right combinations of methods, is far from trivial and that much is still unexplored in what is considered to be the most basic analysis of genomic data.

Motivation:Correct prediction of residue–residue contacts in proteins that lack good templates with known structure would take ab initio protein structure prediction a large step forward. The lack of correct contacts, and in particular long-range contacts, is considered the main reason why these methods often fail.

Results: We propose a novel hidden Markov model (HMM)-based method for predicting residue–residue contacts from protein sequences using as training data homologous sequences, predicted secondary structure and a library of local neighborhoods (local descriptors of protein structure). The library consists of recurring structural entities incorporating short-, medium- and long-range interactions and is general enough to reassemble the cores of nearly all proteins in the PDB. The method is tested on an external test set of 606 domains with no significant sequence similarity to the training set as well as 151 domains with SCOP folds not present in the training set. Considering the top 0.2 · L predictions (L=sequence length), our HMMs obtained an accuracy of 22.8% for long-range interactions in new fold targets, and an average accuracy of 28.6% for long-, medium- and short-range contacts. This is a significant performance increase over currently available methods when comparing against results published in the literature.

Availability: http://predictioncenter.org/Services/FragHMMent/

Contact: torgeir.hvidsten@plantphys.umu.se

Supplementary information: Supplementary data are available at Bioinformatics online.

Background

Sequence similarity to characterized proteins provides testable functional hypotheses for less than 50% of the proteins identified by genome sequencing projects. With structural genomics it is believed that structural similarities may give functional hypotheses for many of the remaining proteins.

Methodology/Principal Findings

We provide a systematic analysis of the structure-function relationship in proteins using the novel concept of local descriptors of protein structure. A local descriptor is a small substructure of a protein which includes both short- and long-range interactions. We employ a library of commonly reoccurring local descriptors general enough to assemble most existing protein structures. We then model the relationship between these local shapes and Gene Ontology using rule-based learning. Our IF-THEN rule model offers legible, high resolution descriptions that combine local substructures and is able to discriminate functions even for functionally versatile folds such as the frequently occurring TIM barrel and Rossmann fold. By evaluating the predictive performance of the model, we provide a comprehensive quantification of the structure-function relationship based only on local structure similarity. Our findings are, among others, that conserved structure is a stronger prerequisite for enzymatic activity than for binding specificity, and that structure-based predictions complement sequence-based predictions. The model is capable of generating correct hypotheses, as confirmed by a literature study, even when no significant sequence similarity to characterized proteins exists.

Conclusions/Significance

Our approach offers a new and complete description and quantification of the structure-function relationship in proteins. By demonstrating how our predictions offer higher sensitivity than using global structure, and complement the use of sequence, we show that the presented ideas could advance the development of meta-servers in function prediction.

Background

We address the issue of explaining the presence or absence of phase-specific transcription in budding yeast cultures under different conditions. To this end we use a model-based detector of gene expression periodicity to divide genes into classes depending on their behavior in experiments using different synchronization methods. While computational inference of gene regulatory circuits typically relies on expression similarity (clustering) in order to find classes of potentially co-regulated genes, this method instead takes advantage of known time profile signatures related to the studied process.

Results

We explain the regulatory mechanisms of the inferred periodic classes with cis-regulatory descriptors that combine upstream sequence motifs with experimentally determined binding of transcription factors. By systematic statistical analysis we show that periodic classes are best explained by combinations of descriptors rather than single descriptors, and that different combinations correspond to periodic expression in different classes. We also find evidence for additive regulation in that the combinations of cis-regulatory descriptors associated with genes periodically expressed in fewer conditions are frequently subsets of combinations associated with genes periodically expression in more conditions. Finally, we demonstrate that our approach retrieves combinations that are more specific towards known cell-cycle related regulators than the frequently used clustering approach.

Conclusion

The results illustrate how a model-based approach to expression analysis may be particularly well suited to detect biologically relevant mechanisms. Our new approach makes it possible to provide more refined hypotheses about regulatory mechanisms of the cell cycle and it can easily be adjusted to reveal regulation of other, non-periodic, cellular processes.

Background

We present an approach designed to identify gene regulation patterns using sequence and expression data collected for Saccharomyces cerevisae. Our main goal is to relate the combinations of transcription factor binding sites (also referred to as binding site modules) identified in gene promoters to the expression of these genes. The novel aspects include local expression similarity clustering and an exact IF-THEN rule inference algorithm. We also provide a method of rule generalization to include genes with unknown expression profiles.

Results

We have implemented the proposed framework and tested it on publicly available datasets from yeast S. cerevisae. The testing procedure consists of thorough statistical analyses of the groups of genes matching the rules we infer from expression data against known sets of co-regulated genes. For this purpose we have used published ChIP-Chip data and Gene Ontology annotations. In order to make these tests more objective we compare our results with recently published similar studies.

Conclusion

Results we obtain show that local expression similarity clustering greatly enhances overall quality of the derived rules, both in terms of enrichment of Gene Ontology functional annotation and coherence with ChIP-Chip binding data. Our approach thus provides reliable hypotheses on co-regulation that can be experimentally verified. An important feature of the method is its reliance only on widely accessible sequence and expression data. The same procedure can be easily applied to other microbial organisms.