Luis Guillermo Leal

Biostatistical approaches for the reconstruction of gene co-expression networks based on transcriptomic data

Techniques in molecular biology have permitted the gathering of an extremely large amount of information relating organisms and their genes. The current challenge is assigning a putative function to thousands of genes that have been detected in different organisms. One of the most informative types of genomic data to achieve a better knowledge of protein function is gene expression data. Based on gene expression data and assuming that genes involved in the same function should have a similar or correlated expression pattern, a function can be attributed to those genes with unknown functions when they appear to be linked in a gene co-expression network (GCN). Several tools for the construction of GCNs have been proposed and applied to plant gene expression data. Here, we review recent methodologies used for plant gene expression data and compare the results, advantages and disadvantages in order to help researchers in their choice of a method for the construction of GCNs.

Construction and comparison of gene co-expression networks shows complex plant immune responses

Gene co-expression networks (GCNs) are graphic representations that depict the coordinated transcription of genes in response to certain stimuli. GCNs provide functional annotations of genes whose function is unknown and are further used in studies of translational functional genomics among species. In this work, a methodology for the reconstruction and comparison of GCNs is presented. This approach was applied using gene expression data that were obtained from immunity experiments in Arabidopsis thaliana, rice, soybean, tomato and cassava. After the evaluation of diverse similarity metrics for the GCN reconstruction, we recommended the mutual information coefficient measurement and a clustering coefficient-based method for similarity threshold selection. To compare GCNs, we proposed a multivariate approach based on the Principal Component Analysis (PCA). Branches of plant immunity that were exemplified by each experiment were analyzed in conjunction with the PCA results, suggesting both the robustness and the dynamic nature of the cellular responses. The dynamic of molecular plant responses produced networks with different characteristics that are differentiable using our methodology. The comparison of GCNs from plant pathosystems, showed that in response to similar pathogens plants could activate conserved signaling pathways. The results confirmed that the closeness of GCNs projected on the principal component space is an indicative of similarity among GCNs. This also can be used to understand global patterns of events triggered during plant immune responses.

Longitudinal analysis of maternal serum Follistatin concentration in normal pregnancy and preeclampsia

Follistatin (FST) is a regulator of the biological activity of activin A (Act A), binding and blocking it, which could contribute to the modulation of its pro‐inflammatory activity during pregnancy. We sought to investigate, in this nested case–control study, FST serum levels during normal pregnancy and correlate it with the FST profile in preeclamptic pregnant women, normal pregnant women followed 3 months postpartum and eumenorrheic nonpregnant women throughout the menstrual cycle.

Maternal serum omentin-1 profile is similar in humans and in the rat animal model

Omentin-1 is an adipocytokine with anti-inflammatory activity that has been associated with different metabolic disorders. The aim of this study is to investigate the serum profiles of omentin-1 throughout human and rat pregnancy. Serum omentin-1 levels were determined by ELISA in a prospective cohort study of healthy pregnant women (n=40) during the three trimesters of pregnancy and in twenty healthy non-pregnant women during the follicular and luteal phase of the menstrual cycle. In addition, serum omentin-1 levels were measured in rats during different periods of pregnancy (gestational days 8, 12, 16, 19, and 21) and in an age-matched control (virgin) group of rats (n=12rats/group). Finally, immunohistochemistry was used to demonstrate the presence of omentin-1 protein in human and rat placenta. Omentin-1 immunoreactivity was detected in cytotrophoblasts, syncytiotrophoblasts, sparse Hofbauer cells, and endothelial cells of the stem villi of human placenta. Additionally, it was detected in the labyrinthine trophoblast and yolk sac layer of the rat placenta. Human and rat serum omentin-1 levels were significantly lower in the late gestational period when compared with the non-pregnant women and virgin rats (p<0.05). Serum omentin-1 changes were not significant throughout the gestation in both species (p>0.05). Human serum omentin-1 levels have an inverse relationship with triglyceride levels during pregnancy. Our findings have not determined the exact role of omentin-1 during pregnancy, concerning the metabolic control of triglycerides and other energy sources. Whether omentin-1 decrease implies a regulatory function is still not clear. Further studies are needed to address this issue and determine the role of omentin-1 in metabolic adaptations during normal human and rat pregnancy.

Identification of immunity-related genes in Arabidopsis and cassava using genomic data.

Recent advances in genomic and post-genomic technologies have provided the opportunity to generate a previously unimaginable amount of information. However, biological knowledge is still needed to improve the understanding of complex mechanisms such as plant immune responses. Better knowledge of this process could improve crop production and management. Here, we used holistic analysis to combine our own microarray and RNA-seq data with public genomic data from Arabidopsis and cassava in order to acquire biological knowledge about the relationships between proteins encoded by immunity-related genes (IRGs) and other genes. This approach was based on a kernel method adapted for the construction of gene networks. The obtained results allowed us to propose a list of new IRGs. A putative function in the immunity pathway was predicted for the new IRGs. The analysis of networks revealed that our predicted IRGs are either well documented or recognized in previous co-expression studies. In addition to robust relationships between IRGs, there is evidence suggesting that other cellular processes may be also strongly related to immunity.

Maternal Serum Meteorin Levels and the Risk of Preeclampsia.

Background Meteorin (METRN) is a recently described neutrophic factor with angiogenic properties. This is a nested case-control study in a longitudinal cohort study that describes the serum profile of METRN during different periods of gestation in healthy and preeclamptic pregnant women. Moreover, we explore the possible application of METRN as a biomarker. Methods and Findings Serum METRN was measured by ELISA in a longitudinal prospective cohort study in 37 healthy pregnant women, 16 mild preeclamptic women, and 20 healthy non-pregnant women during the menstrual cycle with the aim of assessing serum METRN levels and its correlations with other metabolic parameters. Immunostaining for METRN protein was performed in placenta. A multivariate logistic regression model was proposed and a classifier model was formulated for predicting preeclampsia in early and middle pregnancy. The performance in classification was evaluated using measures such as sensitivity, specificity, and the receiver operating characteristic (ROC) curve. In healthy pregnant women, serum METRN levels were significantly elevated in early pregnancy compared to middle and late pregnancy. METRN levels are significantly lower only in early pregnancy in preeclamptic women when compared to healthy pregnant women. Decision trees that did not include METRN levels in the first trimester had a reduced sensitivity of 56% in the detection of preeclamptic women, compared to a sensitivity of 69% when METRN was included. Conclusions The joint measurements of circulating METRN levels in the first trimester and systolic blood pressure and weight in the second trimester significantly increase the probabilities of predicting preeclampsia.

Construction and Comparison of Gene Co-expression Networks Based on Immunity Microarray Data from Arabidopsis, Rice, Soybean, Tomato and Cassava

A big challenge in gene expression data analyses is to reveal the coordinated expression of different genes. Gene co-expression networks (GCNs) are graphic representations where nodes symbolize genes while edges reconstruct the coordinated transcription of genes to certain external stimuli. In this paper, an enhanced novel methodology for construction and comparison of GCNs is proposed. Microarray datasets from pathogen infected plants (Arabidopsis, rice, soybean, tomato and cassava) were used. Initially, similarity metrics that find linear and non-linear correlations between gene expression profiles were evaluated. A similarity threshold was chosen and GCNs were constructed. Afterwards, GCNs were characterized by graph variables and a principal component analysis on these variables was applied to differentiate them. The results allowed the discovery of topologically and non-topologically similar networks among species. Potentially conserved biological processes, like those related to immunity in plants could be studied from this work.

THRESHOLD SELECTION BASED ON RANDOM MATRIX THEORY FOR GENE CO-EXPRESSION NETWORK

 ABSTRACT: Random Matrix Theory (RMT) methods for threshold selection had only been applied in a very low number of studies aiming the construction of Gene Co-expression Networks (GCN) and several open questions remained, especially regarding the general applicability regardless the diverse data structure of gene expression data sets. Moreover, no clear methodology to follow at each step was available. Here, we show, that RMT methodology is, in fact, capable to differentiate Gaussian Orthogonal Ensemble (GOE) from Gaussian Diagonal Ensemble (GDE) structure for a great number of simulated data sets and that results are similar to those obtained with the reference method of clustering coefficient.

Algorithm for the Construction of a Global Enzymatic Network to be used for Gene Network Reconstruction

Relationships between genes are best represented using networks constructed from information of different types, with metabolic information being the most valuable and widely used for genetic network reconstruction. Other types of information are usually also available, and it would be desirable to systematically include them in algorithms for network reconstruction. Here, we present an algorithm to construct a global metabolic network that uses all available enzymatic and metabolic information about the organism. We construct a global enzymatic network (GEN) with a total of 4226 nodes (EC numbers) and 42723 edges representing all known metabolic reactions. As an example we use microarray data for Arabidopsis thaliana and combine it with the metabolic network constructing a final gene interaction network for this organism with 8212 nodes (genes) and 4606,901 edges. All scripts are available to be used for any organism for which genomic data is available.