Skip to main content
NCBI home page
Evolutionary Bioinformatics Online logoLink to Evolutionary Bioinformatics Online
. 2015 Oct 22;11:231–238. doi: 10.4137/EBO.S25585

In Silico Gene Regulatory Network of the Maurer’s Cleft Pathway in Plasmodium falciparum

Itunuoluwa Isewon 1,2,✉, Jelili Oyelade 1,2, Benedikt Brors 3, Ezekiel Adebiyi 1,2,3
PMCID: PMC4620995  PMID: 26526876

Introduction

The existence of multiple-stained structures known to be Maurer’s clefts (MCs) in the cytosol of erythrocytes derived from patients with Plasmodium falciparum malaria using light microscopy was first detected by George Maurer.1–3 For over a century, the functions of the MCs remained obscure.4 In the last decade, scientists have been interested in the mediatory role the MCs play in protein transportation from the cytoplasm of P. falciparum to the surface of the erythrocyte of the host, which they consider an unconventional and intriguing biological procedure.5,6

Upon evasion, P. falciparum totally recasts the infected human red blood cells to access nutrients and to elude the immune system.7 The erythrocytes have no secretory systems that the parasite can subvert. As such, it evolves trafficking pathways for the export of macromolecules.8 More than 10% of all the parasite’s proteins (>400 proteins) are smuggled to the cytosol of its host.7,9 The transported proteins induce extreme structural adjustment to the red blood cell, such as increasing the toughness of the cell of its host and the development of lumps. This in turn makes the affected cells adhere to the endothelium, hence leading to the development of malaria.8,10,11 Reports also show that MCs are created at various portions in the life cycle of the parasite, and this affirms their relevance to parasite development and survival.4

Computationally, only about one-third of the transcription factors (TFs) expected of the genome of the size of P. falciparum are known, while the knowledge of the corresponding binding sites is lacking. A few works have experimentally validated motifs in the parasite, but their binding proteins are unknown. Others have identified some transcription-associated proteins, but very little is known about their binding sites.12–15 This study is an attempt to identify novel transcriptional regulators (master activators/inhibitors) of the MC pathway in P. falciparum by predicting its gene regulatory network (GRN) from gene expression data.

Gene regulatory networks

Gene networks depict transcriptional regulation, ie, how genes respond to signals.16 GRNs simply show how and when genes in a cell interact with each other. GRNs expressly depict the causality of developmental processes.17 Several methods exist for the modeling and simulation of GRNs from high throughput experimental data, such as expression profiles, genomic sequences, and TF-binding site assays. These include, but are not limited to, Boolean models,18,19 Petri net models,20,21 neural network models,22,23 ordinary differential equation models,24,25 Bayesian network (BN) models,24 relevance/association networks,26 and graphical Gaussian models (GGMs).27–29

A commonly used computational method in modeling GRNs is to compute the Pearson correlation coefficients (r) between all possible combinations of the genes of interest. Any of such combinations having Pearson correlation coefficients above a predefined threshold is considered to be significant, ie, exhibit useful relationships, influence, or dependence.27 Such networks are referred to as relevance/association networks. It is important to note that they may lead to vague results, particularly if it is a dense system of connections.30

BNs are classified as probabilistic graphical models.31 BNs are also a type of directed acyclic graphs that are other types of graphical model structures. BNs and its extensions have been used extensively to predict regulatory networks.32–38 However, there are drawbacks with using BNs for modeling GRNs. Some drawbacks include the difficulty in dealing with time series data, ie, higher number of genes compared to sample size, and inability to capture feedback regulation/loops that are normal in real-life gene networks and are nondeterministic polynomial-time hard learning problems.39–41 The implementation of dynamic BNs also has some drawbacks, including excessive computing time, result precision largely dependent on quantity of genes, and requirement of more prior information of the transcription regulation.5,41–43

GGM is a preferred alternative. It uses partial correlation (pcor) to ensure a rich appraisal of a direct relationship between any combinations of genes.44,45 Unlike Pearson correlation, pcor of two genes calculates the amount of correlation left after eliminating the influence of other genes. Previous works have established that GGMs are very effective in inferring conditional dependency and modeling interactions among genes.27,45–49 Earlier, predicting network interactions using GGM had been limited to data sets with fewer genes due to the fact that gene expression experiments had few samples (n), which are usually very less compared to the amount of genes (P).45,46 This scenario is referred to as small sampling problem, ie, P ≫ n cannot be accommodated by standard GGM.49,50 One possibility of resolving this problem is by using the regularized GGM approach proposed by Schäfer and Strimmer.48 This approach is referred to as the shrinkage approach that is suited to cases where the number of genes is slightly larger than the number of samples.51,52 In order to introduce directionality into the network, as GGMs are essentially undirected networks displaying only direct linear relationships, Opgen-Rhein and Strimmer53 introduced an extension of the shrinkage approach to graphical Gaussian modeling, which they referred to as causation networks. We have used this regularized directed GGM, as implemented in the GeneNet R package, to build a gene network for MC pathway of P. falciparum based on gene expression profiles.

Materials and Methods

Microarray data

Gene expression data with accession number GSE24416 were downloaded from the Gene Expression Omnibus (GEO) database (www.ncbi.nlm.nih.gov/geo/, accessed 11/02/2015). The experiment was done by Foth et al.54 The 24 time point samples were harvested from a tightly synchronous 6.5-L biofermenter culture of P. falciparum (Dd2) at two-hour intervals during one entire (48 hours) intraerythrocytic developmental cycle and compared against a 3D7 RNA reference pool.

MC pathway proteins of P. falciparum

Genes that belong to the MCs’ pathway were downloaded from PlasmoDB version 13.0 (http://plasmodb.org, accessed 11/02/2015). Nine pathways concerned with the MCs, based on the classification from Hagai Ginsburg’s Malaria Parasite Metabolic Pathways database (at http://mpmp.huji.ac.il/), were used for this study. They include “Biogenesis of Maurer’s clefts” – 3 genes, “Characteristics of Plasmodium falciparum export proteins that remodel infected erythrocytes” – 41 genes, “Established and putative Maurer’s clefts proteins” – 134 genes, “Exported parasite proteins associated with Maurer’s clefts” – 11 genes, “Genes coding for protein traffic related proteins” – 61 genes, “Maurer’s clefts” – 7 genes, “Properties of proteins exported to erythrocyte” – 39 genes, “Subcellular location of exported proteins” – 43 genes, and “Subcellular localization of proteins involved in invasion” – 81 genes. Expression values were found for 319 of them from the gene expression data set used. The list of genes used for this study with their gene IDs is provided in Supplementary File S01.

Network construction

All analyses were done in R version 3.1.1 (www.R-project.org), Bioconductor version 2.14,55 and RStudio version 0.98.978.56 The following R packages were used for the network modeling: GEOquery version 2.30.1 to load the gene expression data from GEO database directly into R57; GeneNet version 1.2.12 to model the GRN58; Graphviz version 2.38,59 dnet version 1.0.7,60 venerable version 3.0,61 and igraph version 0.7.162 for network visualization; and BCRANK version 1.26.0.63 The detailed computational pipeline used for the analyses is presented in Supplementary File S02. The R source code for the analyses is provided in Supplementary File S03.

Results

The dynamic shrinkage approach as implemented in the GeneNet58 R package was used to estimate the pcor matrix of all possible edges. The resulting pcor network had 50,721 edges connecting 319 nodes with pcor values ranging from −0.20 to +0.213. The density function, distribution function, and local false discovery rate (FDR) values are shown in Figure 1.

Figure 1.

Figure 1

Open in a new tab

Plot of the density function, distribution function, and local FDR values. The null distribution is depicted by the dashed line; it follows a normal distribution with 0 mean and a standard deviation of 0.05. The solid line signifies the alternative distribution. The empirical distribution (indicated by the histogram) is composed of the null distribution and the alternative distribution.

Our approach favored the edges with lower pcor values. This is because pcors having the lowest absolute values mean they have the largest amount of effects after other genes have been removed. We extracted edges whose |pcor| is ≥0.04. This resulted in 1160 edges out of 50,721 edges. The Pearson correlation coefficient (r) was estimated for these edges and used to further prune the network by eliminating edges with −0.60 ≤ r ≤ 0.60 as any r close to 0 indicates independence. This left us with 178 edges. We then selected the top 100 edges (sorted according to the absolute strength of the correlation from the strongest to the weakest), representing ~0.2% of all possible edges as significant edges. The resulting GRN of the MCs’ pathway is shown in Figure 2. The complete significant edge list with their pcor coefficients, Pearson correlation coefficients, P-values, q-values, and FDR values is presented in Supplementary File S04.

Figure 2.

Figure 2

Open in a new tab

The resulting GRN of the MCs’ pathway in P. falciparum. The straight lines represent activation relationship (positive correlation), while the dotted lines represent inhibitory relationship (negative correlation). The intensity of the lines represents the amount of regulation. The red nodes are the hubs – master activators/repressors – while the yellow nodes are their predicted target genes.

From this network, we estimated the out-degree of the participating nodes to identify hubs. The hubs are major players in the network (master activation/master inhibitors). Our hypothesis here is that these genes are possibly playing a key role in the regulation of the pathway, ie, the protein trafficking activities of the parasite, and should be further investigated. From our analysis, four genes had the highest number of direct connections in the network: PFL0780w, PFB0340c, MAL7P1.92, and PF08_0002. We extracted the subnet-works of the genes. Our hypothesis is that these genes are the possible target genes of the respective hub genes. The subnet-works are shown in Figure 3A–D.

Figure 3.

Figure 3

Open in a new tab

Subnet for the master activators/repressors with their target genes. The red nodes are the predicted TFs, while the blue nodes are the predicted target genes that share an overrepresented motif in their promoter region (predicted binding site) and the pink nodes are predicted target genes that do not share the motif but have a direct interaction from the predicted regulatory network in Figure 2: (A) subnetwork for PFL0780w, (B) subnetwork for PFB0340c, (C) subnetwork for MAL7P1.92, and (D) subnetwork for PF08_0002.

Predicting binding sites

We extracted 1000 bp upstream of transcription start site for the target genes identified for each master activator from PlasmoDB in FASTA format. The sequences in each group were ranked (arranged) according to the strength of correlation with their master activator. We then used the methods implemented in the R package BCRANK to find the overrepresented motifs in the upstream sequences. BCRANK is useful for predicting binding site consensus for ranked DNA sequences. See the package manual for more details on the BCRANK algorithm.52 A number of predictions resulted from the running of this algorithm, and the first motif was selected as the candidate motif as it is the most statistically significant motif. Details on the resulting motifs for each set are provided in Supplementary File S05.

Discussion

Our results show four major players that we hypothesized to play important roles in the regulation of activities in the MCs. The first is PFL0780w; from our network, we predicted the following target genes: PFD0955w, PFE1605w, MAL7P1.87, PFI1745c, and PF11_0351. The first two have the motif ACATGTAA overrepresented in their upstream sequence that we predicted to be the binding site of the TF. PFL0780w was reported by the ontology-based pattern identification database to be involved in translational regulator activity and tRNA metabolic process, which is in line with our predictions. From life cycle expression data (combined from several microarray experiments available in PlasmoDB version 13.0), its expression pattern is increasing from the early ring to early trophozoite stage (ie, from 0 to ~25 hpi) and then it reduces; this explains the fact that it is truly an important protein in the survival strategy mechanisms of the parasite. Apart from the MCs, PFL0780w participates in other pathways, such as glycolysis, fatty acid synthesis, and mitochondria electron flow. The second is PFB0340c with the following predicted target genes: PF11_0202, PF14_0758, PF10_0013, PF11_0225, and PF13_0074. The first two genes have the motif TCTTATTTT as the predicted binding site. PFB0340c also known as SERA5 is involved in the regulation of immune response, proteolysis, and immunoglobulin production. SERA5 is predominantly expressed among the SERA multigene family of P. falciparum (which is one of the blood stage malaria vaccine candidates), and the acquired antibody titers correlate with the serum inhibition of the parasite growth.64 Life cycle expression data show that PFB0340c is expressed late in the erythrocytic cycle, ie, from late trophozoite to early schizont stage (~30–35 hpi). The third is MAL7P1.92, and we predicted PFD1160w, PFB0680w, PF11_0168, PF08_0036, and PF14_0529 as its target genes. The last two genes have the motif ATGCCTTTTT overrepresented in their upstream sequence. MAL7P1.92 also known as cysteine repeat modular protein 2 (CRMP2) is localized in parasite plasma membrane during schizont stage and in MC during early trophozoite stage. The fourth is PF08_0002; MAL7P1.164, PF10_0352, PFC1090w, and PF08_0110 are its predicted target genes. The last two genes have the motif TAATAATTTA as their predicted binding site. PF08_0002 is surface-associated interspersed protein 8.2 (SURFIN 8.2). It is upregulated in the early trophozoite stage. Apart from the MC pathway, PF08_0002 is also involved in the interactions between modified host cell membrane and endothelial cell pathway. Table 1 presents more details about each of the four predicted TFs. For each predicted TF, its predicted target genes are listed with their gene names, product description, annotated gene ontology (GO) function term, and annotated GO process term. The enrichment analysis map65,66 for these predicted TFs and their target genes is presented in Figure 4. The following GO categories were found to be significantly overrepresented: protein transport, membrane coat, protein localization, coated membrane, and establishment of protein localization.

Table 1.

Full details of the predicted TFs and their target genes.

TRANSCRIPTION FACTOR TARGET GENES GENE NAME PRODUCT DESCRIPTION ANNOTATED GO FUNCTION TERM ANNOTATED GO PROCESS TERM
PFL0780w Glycerol-3-phosphate dehydrogenase, putative NAD binding, glycerol-3-phosphate dehydrogenase [NAD+] activity, protein homodimerization activity Carbohydrate metabolic process, glycerol-3-phosphate catabolic process, glycerol-3-phosphate metabolic process, oxidation-reduction process
PFD0955w Pf34 Apical merozoite protein (Pf34)
PFE1605w LyMP Lysine-rich membrane-associated PHISTb protein (LyMP) Protein binding
MAL7P1.87 Conserved Plasmodium protein, unknown function ATP binding, actin binding, calmodulin binding, motor activity
PFI1745c Early transcribed membrane protein
PF11_0351 HSP70–3 Heat shock protein 70 (HSP70–3) ATP binding, unfolded protein binding Protein folding, response to heat, response to unfolded protein
PFB0340c SERA5 Serine repeat antigen 5 (SERA5) Cysteine-type peptidase activity Immunoglobulin production, proteolysis, regulation of immune response
PF11_0202 AP-4 complex subunit mu, putative Protein binding Intracellular protein transport, vesicle-mediated transport
PF14_0758 PTP3 EMP1-trafficking protein (PTP3)
PF10_0013 Plasmodium exported protein (hyp12), unknown function (PfJ13)
PF11_0225 GCN20 Protein GCN20 (GCN20) ATP binding, ATPase activity, coupled to transmembrane movement of substances Regulation of translation, transport
PF13_0074 SURF13.1 Surface-associated interspersed protein 13.1 (SURFIN 13.1), pseudo-gene (SURF13.1)
MAL7P1.92 CRMP2 Cysteine repeat modular protein 2 (CRMP2) Protein binding Intracellular receptor mediated signalling pathway, intracellular transport
PFD1160w SURF4.2 Surface-associated interspersed protein 4.2 (SURFIN 4.2) (SURF4.2)
PFB0680w RON6 Rhoptry neck protein 6 (RON6) Entry into host
PF11_0168 RON4 Rhoptry neck protein 4 (RON4)
PF08_0036 SEC23 Protein transport protein SEC23 (SEC23) Zinc ion binding ER to Golgi vesicle-mediated transport, intracellular protein transport
PF14_0529 AP-1 complex subunit gamma, putative Binding, protein transporter activity Intracellular protein transport, vesicle-mediated transport
PF08_0002 SURF8.2 Surface-associated interspersed protein 8.2 (SURFIN 8.2) (SURF8.2) Binding, protein transporter activity
MAL7P1.164 AP-4 complex subunit beta, putative Binding, protein transporter activity Intracellular protein transport, vesicle-mediated transport
PF10_0352 MSP11 Merozoite surface protein (MSP11)
PFC1090w EPF4 Exported protein family 4 (EPF4)
PF08_0110 RAB18 Ras-related protein Rab-18 (RAB18) GTP binding, GTPase activity GTP catabolic process, intracellular protein transport, nucleocytoplasmic transport, regulation of vesicle- mediated transport, small GTPase mediated signal transduction
Open in a new tab

Figure 4.

Figure 4

Open in a new tab

Enrichment analysis map of all predicted TFs and their target genes for the MCs’ pathway at a significant level of P = 0.001 and q = 0.05. The enrichment map gives a graphical representation of the GO terms that were found to be significantly overrepresented in the context of the GO hierarchy.

Conclusion

We have successfully predicted the GRN of a very important pathway in the deadly malaria parasite P. falciparum. A number of predictions have been made from our analyses that should be further validated experimentally. Our results have also confirmed findings from previous studies and yielded new information, which shed more light and help improve our understanding of the genetic regulation activities of P. falciparum in our quest to fully overcome the malaria endemic. The knowledge of the predicted TFs, their target genes, and their predicted binding sites will indeed increase our understanding of the mechanism that underlay the tight transcriptional control of P. falciparum and also help in the rational design of promoters to ease the study of essential genes of the parasite.

Supplementary Material

Supplementary File S01. Gene_list.txt.

Supplementary File S02. Computational pipeline.pdf.

Supplementary File S03. MC-GRN_sourcecode.R.

Supplementary File S04. Filtered_network.xlsx.

Supplementary File S05. Motifs.txt.

EBO-11-2015-231-s001.zip (364.7KB, zip)

Acknowledgments

The authors would like to acknowledge the contribution of the reviewers in improving the quality of this article.

Footnotes

ACADEMIC EDITOR: Jike Cui, Associate Editor

PEER REVIEW: Three peer reviewers contributed to the peer review report. Reviewers’ reports totaled 2,269 words, excluding any confidential comments to the academic editor.

FUNDING: This work was supported by study grants from the German Academic Exchange Service (DAAD), Covenant University, and German Cancer Research Centre (DKFZ). The authors confirm that the funder had no influence over the study design, content of the article, or selection of this journal.

COMPETING INTERESTS: Authors disclose no potential conflicts of interest.

Paper subject to independent expert blind peer review. All editorial decisions made by independent academic editor. Upon submission manuscript was subject to anti-plagiarism scanning. Prior to publication all authors have given signed confirmation of agreement to article publication and compliance with all applicable ethical and legal requirements, including the accuracy of author and contributor information, disclosure of competing interests and funding sources, compliance with ethical requirements relating to human and animal study participants, and compliance with any copyright requirements of third parties. This journal is a member of the Committee on Publication Ethics (COPE).

Author Contributions

All authors contributed to this work. Original idea and conception: II, JO, BB, EA. Implementation of research methods: II. Writing and revision of the manuscript: II, JO. All authors reviewed and approved the final manuscript.

REFERENCES

  • 1.Maurer G. Die malaria perniciosa. Zentralbl Bakteriol Parasitenkunde. 1902;23:695–719. [Google Scholar]
  • 2.Trager W, Rudzinska MA, Bradbury PC. The fine structure of Plasmodium falciparum and its host erythrocyte in natural malarial infections in man. Bull World Health Organ. 1966;35(3):883–5. [PMC free article] [PubMed] [Google Scholar]
  • 3.Langreth SG, Jensen JB, Reese RT, Trager W. Fine structure of human malaria in vitro. J Protozool. 1978;25(4):443–52. doi: 10.1111/j.1550-7408.1978.tb04167.x. [DOI] [PubMed] [Google Scholar]
  • 4.Lanzer M, Wickert H, Krohne G, Vincensini L, Braun Breton C. Maurer’s clefts: a novel multi-functional organelle in the cytoplasm of Plasmodium falciparum-infected erythrocytes. Int J Parasitol. 2006;36(1):23–36. doi: 10.1016/j.ijpara.2005.10.001. [DOI] [PubMed] [Google Scholar]
  • 5.Ewejobi IM, Bulashevska S, Brors B, Adebiyi EF. Proceeding of the EMBnet-RIBio. Vol. 16. Cancun, Mexico: 2009. In-silico prediction of the genetic regulatory interactions in Maurer’s cleft pathway of Plasmodium falciparum. EMBnet.journal. [Google Scholar]
  • 6.Frischknecht F, Lanzer M. The Plasmodium falciparum Maurer’s clefts in 3D. Mol Microbiol. 2008;67(4):687–91. doi: 10.1111/j.1365-2958.2007.06095.x. [DOI] [PubMed] [Google Scholar]
  • 7.Mundwiler-Pachlatko E, Beck HP. Maurer’s clefts, the enigma of Plasmodium falciparum. Proc Natl Acad Sci U S A. 2013;110(50):19987–94. doi: 10.1073/pnas.1309247110. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Sam-Yellowe TY. The role of the Maurer’s clefts in protein transport in Plasmodium falciparum. Trends Parasitol. 2009;25(6):277–84. doi: 10.1016/j.pt.2009.03.009. [DOI] [PubMed] [Google Scholar]
  • 9.Miller EA, Beilharz TH, Malkus PN, et al. Multiple cargo binding sites on the COPII subunit Sec24p ensure capture of diverse membrane proteins into transport vesicles. Cell. 2003;114(4):497–509. doi: 10.1016/s0092-8674(03)00609-3. [DOI] [PubMed] [Google Scholar]
  • 10.Kirchgatter K, Del Portillo HA. Clinical and molecular aspects of severe malaria. An Acad Bras Cienc. 2005;77(3):455–75. doi: 10.1590/s0001-37652005000300008. [DOI] [PubMed] [Google Scholar]
  • 11.Haldar K, Mohandas N, Bhattacharjee S, et al. Trafficking and the tubulovesicular membrane network. In: Sherman IW, editor. Molecular Approaches to Malaria. 1st ed. Washington, DC: ASM Press; 2005. pp. 253–71. [Google Scholar]
  • 12.Young JA, Johnson JR, Benner C, et al. In silico discovery of transcription regulatory elements in Plasmodium falciparum. BMC Bioinformatics. 2008;9:70. doi: 10.1186/1471-2164-9-70. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Iengar P, Joshi NV. Identification of putative regulatory motifs in the up-stream regions of co-expressed functional groups of genes in Plasmodium falciparum. BMC Genomics. 2009;10:18. doi: 10.1186/1471-2164-10-18. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Coulson RMR, Hall N, Ouzounis CA. Comparative genomics of transcriptional control in the human malaria parasite Plasmodium falciparum. Genome Res. 2004;14:1548–54. doi: 10.1101/gr.2218604. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Bischoff E, Vaquero C. In silico and biological survey of transcription-associated proteins implicated in the transcriptional machinery during the erythrocyctic development of Plasmodium falciparum. BMC Genomics. 2010;11:34. doi: 10.1186/1471-2164-11-34. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Schlitt T, Brazma A. Current approaches to gene regulatory network modeling. BMC Bioinformatics. 2007;8(suppl 6):S9. doi: 10.1186/1471-2105-8-S6-S9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Davidson E, Levin M. Gene regulatory networks. Proc Natl Acad Sci U S A. 2005;102(14):4935. doi: 10.1073/pnas.0502024102. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Albert R. Boolean modeling of genetic regulatory networks. In: Ben-Naim E, Frauenfelder H, Toroczkai Z, editors. Complex Networks. Berlin: Springer; 2004. pp. 459–81. [Google Scholar]
  • 19.Lovrics A, Gao Y, Juhász B, et al. Boolean modelling reveals new regulatory connections between transcription factors orchestrating the development of the ventral spinal cord. PloS One. 2013;9(11):e111430. doi: 10.1371/journal.pone.0111430. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Durzinsky M, Wagler A, Marvan W. Reconstruction of extended petri nets from time series data and its application to signal transduction and to gene regulatory networks. BMC Syst Biol. 2011;5:113. doi: 10.1186/1752-0509-5-113. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Marwan W, Rohr C, Heiner M. Petri nets in snoopy: a unifying framework for the graphical display, computational modelling, and simulation of bacterial regulatory networks. In: van Helden J, Toussaint A, Thieffry D, editors. Bacterial Molecular Networks. New York: Springer; 2012. pp. 409–37. [DOI] [PubMed] [Google Scholar]
  • 22.Xu R, Wunsch IID, Frank R. Inference of genetic regulatory networks with recurrent neural network models using particle swarm optimization. IEEE/ACM Trans Comput Biol Bioinform. 2007;4(4):681–92. doi: 10.1109/TCBB.2007.1057. [DOI] [PubMed] [Google Scholar]
  • 23.Cheng L, Hou ZG, Lin Y, Tan M, Zhang WC, Wu FX. Recurrent neural network for non-smooth convex optimization problems with application to the identification of genetic regulatory networks. IEEE Trans Neural Netw. 2011;22(5):714–26. doi: 10.1109/TNN.2011.2109735. [DOI] [PubMed] [Google Scholar]
  • 24.Li Z, Li P, Krishnan A, Liu J. Large-scale dynamic gene regulatory network inference combining differential equation models with local dynamic Bayesian network analysis. Bioinformatics. 2011;27(19):2686–91. doi: 10.1093/bioinformatics/btr454. [DOI] [PubMed] [Google Scholar]
  • 25.Cao J, Qi X, Zhao H. Modeling gene regulation networks using ordinary differential equations. In: Wang J, Tan AC, Tian T, editors. Next Generation Microarray Bioinformatics. Humana Press; Totowa, NJ: 2012. pp. 185–97. [DOI] [PubMed] [Google Scholar]
  • 26.Butte AS, Kohane IS. Relevance networks: a first step toward finding genetic regulatory networks within microarray data. In: Parmigiani G, Garett ES, Irizarry RA, Zeger SL, editors. The Analysis of Gene Expression Data. New York: Springer; 2003. pp. 428–46. [Google Scholar]
  • 27.Ma S, Gong Q, Bohnert HJ. An Arabidopsis gene network based on the graphical Gaussian model. Genome Res. 2007;17(11):1614–25. doi: 10.1101/gr.6911207. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Ingkasuwan P, Netrphan S, Prasitwattanaseree S, et al. Inferring transcriptional gene regulation network of starch metabolism in Arabidopsis thaliana leaves using graphical Gaussian model. BMC Syst Biol. 2012;6:100. doi: 10.1186/1752-0509-6-100. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Krouk G, Lingeman J, Colon AM, Coruzzi G, Shasha D. Gene regulatory networks in plants: learning causality from time and perturbation. Genome Biol. 2013;14(6):123. doi: 10.1186/gb-2013-14-6-123. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Brazhnik P, de la Fuente A, Mendes P. Gene networks: how to put the function in genomics. Trends Biotechnol. 2002;20(11):467–72. doi: 10.1016/s0167-7799(02)02053-x. [DOI] [PubMed] [Google Scholar]
  • 31.Harel A, Kenett RS, Ruggeri F. Modeling web usability diagnostics on the basis of usage statistics. In: Jank W, Shmueli G, editors. Statistical Methods in e-Commerce Research. Hoboken, NJ: John Wiley & Sons, Inc; 2008. pp. 131–72. [Google Scholar]
  • 32.Gallo CA, Carballido JA, Ponzoni I. Inference of gene regulatory networks based on association rules. In: Elloumi M, Zomaya AY, editors. Biological Knowledge Discovery Handbook: Preprocessing, Mining, and Postprocessing of Biological Data. Hoboken, NJ: John Wiley & Sons, Inc; 2013. pp. 803–40. [Google Scholar]
  • 33.Armaanzas R, Inza I, Larraaga P. Detecting reliable gene interactions by a hierarchy of Bayesian network classifiers. Comput Methods Programs Biomed. 2008;91(2):110–21. doi: 10.1016/j.cmpb.2008.02.010. [DOI] [PubMed] [Google Scholar]
  • 34.Djebbari A, Quackenbush J. Seeded Bayesian networks: constructing genetic networks from microarray data. BMC Syst Biol. 2008;2:57. doi: 10.1186/1752-0509-2-57. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Dojer N, Gambin A, Mizera A, Wilczyński B, Tiuryn J. Applying dynamic Bayesian networks to perturbed gene expression data. BMC Bioinformatics. 2006;7:249. doi: 10.1186/1471-2105-7-249. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Kim S, Imoto S, Miyano S. Dynamic Bayesian network and nonparametric regression for nonlinear modeling of gene networks from time series gene expression data. Biosystems. 2004;75(1–3):57–65. doi: 10.1016/j.biosystems.2004.03.004. [DOI] [PubMed] [Google Scholar]
  • 37.Wang M, Chen Z, Cloutier S. A hybrid Bayesian network learning method for constructing gene networks. Comput Biol Chem. 2007;31(5–6):361–72. doi: 10.1016/j.compbiolchem.2007.08.005. [DOI] [PubMed] [Google Scholar]
  • 38.Werhli AV, Husmeier D. Reconstructing gene regulatory networks with Bayesian networks by combining expression data with multiple sources of prior knowledge. Stat Appl Genet Mol Biol. 2007;6(1):15. doi: 10.2202/1544-6115.1282. [DOI] [PubMed] [Google Scholar]
  • 39.Lee WP, Yang KC. A clustering-based approach for inferring recurrent neural networks as gene regulatory networks. Neurocomputing. 2008;71(4–6):600–10. [Google Scholar]
  • 40.Neapolitan RE. Learning Bayesian Networks. Upper Saddle River, NJ: Prentice Hall; 2003. [Google Scholar]
  • 41.Ristevski B. A survey of models for inference of gene regulatory networks. Nonlinear Anal Model Control. 2013;18(4):444–65. [Google Scholar]
  • 42.Zhang Y, Deng Z, Jiang H, Jia P. Inferring gene regulatory networks from multiple data sources via a dynamic Bayesian networks with structural EM. Lect Notes Comput Sci. 2007;4544:204–14. [Google Scholar]
  • 43.Zou M, Conzen SD. A new dynamic Bayesian network (DBN) approach for identifying gene regulatory networks from time course microarray data. Bioinformatics. 2005;21(1):71–9. doi: 10.1093/bioinformatics/bth463. [DOI] [PubMed] [Google Scholar]
  • 44.Whittaker J. Graphical Models in Applied Multivariate Statistics. New York: Wiley; 1990. [Google Scholar]
  • 45.Toh H, Horimoto K. Inference of a genetic network by a combined approach of cluster analysis and graphical Gaussian modeling. Bioinformatics. 2002;18(2):287–97. doi: 10.1093/bioinformatics/18.2.287. [DOI] [PubMed] [Google Scholar]
  • 46.Kishino H, Waddell PJ. Correspondence analysis of genes and tissue types and finding genetic links from microarray data. Genome Inform Ser Workshop Genome Inform. 2000;11:83–95. [PubMed] [Google Scholar]
  • 47.Magwene PM, Kim J. Estimating genomic coexpression networks using first-order conditional independence. Genome Biol. 2004;5(12):R100. doi: 10.1186/gb-2004-5-12-r100. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48.Schäfer J, Strimmer K. Learning large-scale graphical Gaussian models from genomic data. In: Mendes J, editor. Proceedings of “Science of Complex Networks: From Biology to the Internet and WWW” (CNET 2004) Aveiro, Portugal: The American Institute of Physics; 2005. pp. 263–76. [Google Scholar]
  • 49.Wille A, Buhlmann P. Low-order conditional independence graphs for inferring genetic networks. Stat Appl Genet Mol Biol. 2006;5:1. doi: 10.2202/1544-6115.1170. [DOI] [PubMed] [Google Scholar]
  • 50.Schäfer J, Strimmer K. An empirical Bayes approach to inferring large-scale gene association networks. Bioinformatics. 2005;21(6):754–64. doi: 10.1093/bioinformatics/bti062. [DOI] [PubMed] [Google Scholar]
  • 51.Schäfer J Strimmer K. A shrinkage approach to large-scale covariance matrix estimation and implications for functional genomics. Stat Appl Genet Mol Biol. 2005;4:Article32. doi: 10.2202/1544-6115.1175. [DOI] [PubMed] [Google Scholar]
  • 52.Schäfer J Opgen-Rhein R, K Strimmer. Reverse engineering genetic networks using the “GeneNet” package. R News. 2006;6(5):50–3. [Google Scholar]
  • 53.Opgen-Rhein R, Strimmer K. From correlation to causation networks: a simple approximate learning algorithm and its application to high-dimensional plant gene expression data. BMC Syst Biol. 2007;1:37. doi: 10.1186/1752-0509-1-37. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 54.Foth BJ, Zhang N, Chaal BK, Sze SK, Preiser PR, Bozdech Z. Quantitative time-course profiling of parasite and host cell proteins in the human malaria parasite Plasmodium falciparum. Mol Cell Proteomics. 2011;10(8):M110.006411. doi: 10.1074/mcp.M110.006411. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 55.Gentleman RC, Carey VJ, Bates DM, et al. Bioconductor: open software development for computational biology and bioinformatics. Genome Biol. 2004;5(10):R80. doi: 10.1186/gb-2004-5-10-r80. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 56.RStudio . RStudio: Integrated Development Environment for R (Version 0.98.978) [Computer software] Boston, MA: 2014. Available at: http://www.rstudio.org/ [Google Scholar]
  • 57.Davis S, Meltzer PS. GEOquery: a bridge between the gene expression omnibus (GEO) and bioconductor. Bioinformatics. 2007;14:1846–7. doi: 10.1093/bioinformatics/btm254. [DOI] [PubMed] [Google Scholar]
  • 58.Schäfer J Opgen-Rhein R, K Strimmer. GeneNet: Modeling and Inferring Gene Networks. 2015. R package version 1.2.12.
  • 59.Gansner ER, North SC. An open graph visualization system and its applications to software engineering. Software Pract Exp. 2000;30(11):1203–33. [Google Scholar]
  • 60.Fang H, Gough J. The ‘dnet’ approach promotes emerging research on cancer patient survival. Genome Med. 2014;6:64. doi: 10.1186/s13073-014-0064-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 61.Swinton J. Vennerable: Venn and Euler Area-Proportional Diagrams. 2013. R package version 3.0/r82.
  • 62.Csardi G, Nepusz T. The igraph software package for complex network research. Int Complex Syst. 2006;1695(5) [Google Scholar]
  • 63.Ameur A. BCRANK: Predicting Binding Site Consensus from Ranked DNA Sequences. 2010. R package version 1.26.0.
  • 64.Aoki S, Li J, Itagaki S, et al. Serine repeat antigen (SERA5) is predominantly expressed among the SERA multigene family of Plasmodium falciparum, and the acquired antibody titers correlate with serum inhibition of the parasite growth. J Biol Chem. 2002;277(49):47533–40. doi: 10.1074/jbc.M207145200. [DOI] [PubMed] [Google Scholar]
  • 65.Huang DW, Sherman BT, Lempicki RA. Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources. Nat Protoc. 2009;4(1):44–57. doi: 10.1038/nprot.2008.211. [DOI] [PubMed] [Google Scholar]
  • 66.Merico D, Isserlin R, Stueker O, Emili A, Bader GD. Enrichment map: a network-based method for gene-set enrichment visualization and interpretation. PLoS One. 2010;5(11):e13984. doi: 10.1371/journal.pone.0013984. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary File S01. Gene_list.txt.

Supplementary File S02. Computational pipeline.pdf.

Supplementary File S03. MC-GRN_sourcecode.R.

Supplementary File S04. Filtered_network.xlsx.

Supplementary File S05. Motifs.txt.

EBO-11-2015-231-s001.zip (364.7KB, zip)

Articles from Evolutionary Bioinformatics Online are provided here courtesy of SAGE Publications

RESOURCES