Mining DNA microarray data using a novel approach based on graph theory

Basic networks: definition and applications

We define basic networks as the undirected subgraphs with minimal number of units in which the distances (geodesics, minimal path lengths) among a set of selected nodes, which we call seeds, in the original graph are conserved. The additional nodes required to draw the basic network are called connectors. We describe a heuristic strategy to find the basic networks of complex graphs. We also show how the characterization of these networks may help to obtain relevant biological information from highly complex protein-protein interaction data.

Connectivity can be used to identify key genes in DNA microarray data: a study based on gene expression in nasal polyps before and after treatment with glucocorticoids

CONCLUSIONS

The presented analysis of nasal polyposis using connectivity based on the PubGene literature co-citation network demonstrates that this tool can be used to identify key genes in DNA microarray studies of human polygenic diseases.

OBJECTIVES

DNA microarray studies of complex diseases may reveal differential expression of hundreds of genes. According to network theory and studies of yeast cells, genes that are connected with several other genes appear to have key regulatory roles. This study aimed to examine if this principle can be translated to DNA microarray studies of human disease, using nasal polyposis as a base for the analysis.

MATERIALS AND METHODS

The connectivity of differentially expressed genes from a previously described microarray study of nasal polyposis before and after treatment with glucocorticoids was determined. This was done using the literature co-citation network PubGene.

RESULTS

In all, 166 genes were differentially expressed; 39 of these were previously defined as inflammatory and considered important for nasal polyposis. The connectivity of all differentially expressed genes was analysed using the PubGene literature co-citation network. Seventy-four of the 166 genes were connected to other genes. By contrast, the average number of connected genes among 100 sets of 166 randomly chosen genes was 31.5. A small number of the differentially expressed genes were highly connected, while most genes had few or no connections. This indicated a scale-free network. The most connected gene was interleukin-8, an inflammatory gene of known importance for nasal polyposis. Twenty-eight of the 74 connected genes were inflammatory (38%), compared with 11 of the 92 unconnected genes (12%), p < 0.0001. Since most evidence suggests that nasal polyps are inflammatory in their nature, this supports the hypothesis that connected genes have more disease relevance than unconnected genes.

Neuroprotection by endogenous and exogenous PACAP following stroke

We investigated the effects of PACAP treatment, and endogenous PACAP deficiency, on infarct volume, neurological function, and the cerebrocortical transcriptional response in a mouse model of stroke, middle cerebral artery occlusion (MCAO). PACAP-38 administered i.v. or i.c.v. 1 h after MCAO significantly reduced infarct volume, and ameliorated functional motor deficits measured 24 h later in wild-type mice. Infarct volumes and neurological deficits (walking faults) were both greater in PACAP-deficient than in wild-type mice, but treatment with PACAP reduced lesion volume and neurological deficits in PACAP-deficient mice to the same level of improvement as in wild-type mice. A 35,546-clone mouse cDNA microarray was used to investigate cortical transcriptional changes associated with cerebral ischemia in wild-type and PACAP-deficient mice, and with PACAP treatment after MCAO in wild-type mice. 229 known (named) transcripts were increased (228) or decreased (1) in abundance at least 50% following cerebral ischemia in wild-type mice. 49 transcripts were significantly up-regulated only at 1 h post-MCAO (acute response transcripts), 142 were up-regulated only at 24 h post-MCAO (delayed response transcripts) and 37 transcripts were up-regulated at both times (sustained response transcripts). More than half of these are transcripts not previously reported to be altered in ischemia. A larger percentage of genes up-regulated at 24 hr than at 1 hr required endogenous PACAP, suggesting a more prominent role for PACAP in later response to injury than in the initial response. This is consistent with a neuroprotective role for PACAP in late response to injury, i.e., even when administered 1 hr or more after MCAO. Putative injury effector transcripts regulated by PACAP include beta-actin, midline 2, and metallothionein 1. Potential neuroprotective transcripts include several demonstrated to be PACAP-regulated in other contexts. Prominent among these were transcripts encoding the PACAP-regulated gene Ier3, and the neuropeptides enkephalin, substance P (tachykinin 1), and neurotensin.

Improved prediction of critical residues for protein function based on network and phylogenetic analyses

BACKGROUND

Phylogenetic approaches are commonly used to predict which amino acid residues are critical to the function of a given protein. However, such approaches display inherent limitations, such as the requirement for identification of multiple homologues of the protein under consideration. Therefore, complementary or alternative approaches for the prediction of critical residues would be desirable. Network analyses have been used in the modelling of many complex biological systems, but only very recently have they been used to predict critical residues from a protein's three-dimensional structure. Here we compare a couple of phylogenetic approaches to several different network-based methods for the prediction of critical residues, and show that a combination of one phylogenetic method and one network-based method is superior to other methods previously employed.

RESULTS

We associate a network with each member of a set of proteins for which the three-dimensional structure is known and the critical residues have been previously determined experimentally. We show that several network-based centrality measurements (connectivity, 2-connectivity, closeness centrality, betweenness and cluster coefficient) accurately detect residues critical for the protein's function. Phylogenetic approaches render predictions as reliable as the network-based measurements, although, interestingly, the two general approaches tend to predict different sets of critical residues. Hence we propose a hybrid method that is composed of one network-based calculation--the closeness centrality--and one phylogenetic approach--the Conseq server. This hybrid approach predicts critical residues more accurately than the other methods tested here.

CONCLUSION

We show that network analysis can be used to improve the prediction of amino acids critical for protein function, when utilized in combination with phylogenetic approaches. It is proposed that such improvement is due to the complementary nature of these approaches: network-based methods tend to predict as critical those residues that are highly connected and internal (i.e., non-surface), although some surface residues are indeed identified as critical by network analyses; whereas residues chosen by phylogenetic approaches display a lower overall probability of being surface inaccessible.

Bioinformatic methods for integrating whole-genome expression results into cellular networks

Extracting a comprehensive overview from the huge amount of information arising from whole-genome analyses is a significant challenge. This review critically surveys the state of the art methods that are used to connect information from functional genomic studies to biological function. Cluster analysis methods for inferring the correlation between genes are discussed, as are the methods for integrating gene expression information with existing information on biological pathways and the methods that combine cluster analysis with biological information to reconstruct novel biological networks.

Global properties of biological networks

This article discusses the most recent achievements in understanding the biological implications of the small-world and scale-free global topological properties of genetic, proteomic and metabolic networks. Most importantly, these networks are highly clustered and have small node-to-node distances. With their few very connected nodes, which are statistically unlikely to fail under random conditions, the proper functioning of these systems is maintained under external perturbations.

Content-rich biological network constructed by mining PubMed abstracts

BACKGROUND

The integration of the rapidly expanding corpus of information about the genome, transcriptome, and proteome, engendered by powerful technological advances, such as microarrays, and the availability of genomic sequence from multiple species, challenges the grasp and comprehension of the scientific community. Despite the existence of text-mining methods that identify biological relationships based on the textual co-occurrence of gene/protein terms or similarities in abstract texts, knowledge of the underlying molecular connections on a large scale, which is prerequisite to understanding novel biological processes, lags far behind the accumulation of data. While computationally efficient, the co-occurrence-based approaches fail to characterize (e.g., inhibition or stimulation, directionality) biological interactions. Programs with natural language processing (NLP) capability have been created to address these limitations, however, they are in general not readily accessible to the public.

RESULTS

We present a NLP-based text-mining approach, Chilibot, which constructs content-rich relationship networks among biological concepts, genes, proteins, or drugs. Amongst its features, suggestions for new hypotheses can be generated. Lastly, we provide evidence that the connectivity of molecular networks extracted from the biological literature follows the power-law distribution, indicating scale-free topologies consistent with the results of previous experimental analyses.

CONCLUSIONS

Chilibot distills scientific relationships from knowledge available throughout a wide range of biological domains and presents these in a content-rich graphical format, thus integrating general biomedical knowledge with the specialized knowledge and interests of the user. Chilibot http://www.chilibot.net can be accessed free of charge to academic users.

A preprocessing method for inferring genetic interaction from gene expression data using Boolean algorithm

Unknown genetic regulation mechanisms are expected to be discovered by information technology using large amount of biological data especially for gene expression data. In this study, we propose a novel inferring method for genetic interactions that combines our original preprocessing method and the Boolean algorithm. First, the performance of our method was evaluated using artificial data. The results showed that our method was able to infer genetic interactions with high specificity (specificity=0.629). Then, using our method, the genetic interaction was inferred from the experimental time course data collected using microarray on 69 genes of cell cycle for Saccharomyces cerevisiae. Our method estimated about 80% of all genetic interactions in Kyoto Encyclopedia Genes and Genomes (KEGG) for these genes. Furthermore, our method was able to infer several other genetic interactions that are not included in KEGG but whose existence is supported by other biological reports.

Mining gene expression data using a novel approach based on hidden Markov models

In this work we have developed a new framework for microarray gene expression data analysis. This framework is based on hidden Markov models. We have benchmarked the performance of this probability model-based clustering algorithm on several gene expression datasets for which external evaluation criteria were available. The results showed that this approach could produce clusters of quality comparable to two prevalent clustering algorithms, but with the major advantage of determining the number of clusters. We have also applied this algorithm to analyze published data of yeast cell cycle gene expression and found it able to successfully dig out biologically meaningful gene groups. In addition, this algorithm can also find correlation between different functional groups and distinguish between function genes and regulation genes, which is helpful to construct a network describing particular biological associations. Currently, this method is limited to time series data. Supplementary materials are available at http://www.bioinfo.tsinghua.edu.cn/~rich/hmmgep_supp/.

cDNA microarray analysis of changes in gene expression induced by neuronal hypoxia in vitro

We used cDNA microarray gene expression profiling to characterize the transcriptional response to exposure of cultured mouse cerebral cortical neurons to hypoxia for 24 hr. Of 11,200 genes examined, 1,405 (12.5%) were induced or repressed at least 1.5-fold, whereas 26 known genes were induced and 20 known genes were repressed at least 2.5-fold. The most strongly induced genes included genes coding for endoplasmic reticulum proteins (Ero1L/Giig11, Sac1p, Ddit3/Gadd153), proteins involved in ubiquitination (Arih2, P4hb), proteins induced by hypoxia in non-neuronal systems (Gpi1, Aldo1, Anxa2, Hig1), and proteins that might promote cell death (Gas5, Egr1, Ndr1, Vdac2). These findings reinforce the importance of endoplasmic reticulum-based mechanisms and of protein-ubiquitination pathways in the neuronal response to hypoxia.