Development of glomerulosclerotic alterations is a common pathological feature of various progressive kidney diseases. The earliest stages of these different disease entities are characterized by common morphological and functional alterations of the glomeruli, such as glomerular hypertrophy and consecutive development of albuminuria. The present study addressed the question, if such common patterns of morphological and functional glomerular alterations would also find a reflection in common glomerular geneexpressionprofiles. Therefore differential geneexpressionprofiles of samples of isolated kidney glomeruli from two different transgenic mouse models of nephropathy were identified. Microarray experiments were performed in two defined comparable early stages of glomerular alteration. Investigated transgenic murine models of nephropathy consisted of a novel model of diabetes mellitus, transgenic mice expressing a dominant negative glucose-dependent insulinotropic polypeptide receptor (GIPR dn ), bred on the genetic background of the CD1 outbred stock; and growth hormone-transgenic mice (bGH), bred on a NMRI background. Transgenic animals of both models develop glomerular hypertrophy and micro- albuminuria. Pairs of male transgenic mice and their corresponding non-transgenic littermate control animals were investigated in two early comparable stages of glomerular alteration. These stages were defined as the stage of glomerular hypertrophy (stage I), characterized by a significant increase (40 – 60%) of the mean glomerular volume of the transgenic animals compared to the respective controls; and as stage II, the stage of onset of albuminuria. Transgenic animals assigned to stage II also had to display a significant increase of their mean glomerular volumes, as well as an onset of albuminuria, determined by repeated SDS-PAGE based urine analyses of urine samples taken on consecutive time points. Albuminuria was verified by Western blot and ELISA experiments. At both stages transgenic mice displayed a significant increase in numbers of mesangial and endothelial cells per glomerulus, determined by quantitative stereology, while numbers of podocytes per glomerulus remained almost unchanged. Glomerulus isolation was performed according to a modified magnetic isolation procedure, using spherical superparamagnetic beads for perfusion.
data from all 31 children were analyzed. In a first step, the 11 samples of the two cytogenetically homoge- neous subgroups were projected into the space of the first two principal components (PC) of the 6 ribosomal proteins (with specific probe sets; _at) identified as differentially expressed in the subanalysis (S27a, S29, L18, L22, L24, L44; Figure 2). This showed that the risk class assignment was clearly dependent on the riboso- mal expression. The first two PC accurately discrimi- nated the two subgroups (see Supplemental Figure B from URL http://www.kispi.unizh.ch/onkologie). The Figure 2. Gene signature associated with risk groups. Top discriminating genes between 5 TEL-AML1 positive patients classified as standard-risk (SR TEL-AML1) and 6 high-risk patients without distinct cytogenetic alterations (indicated as HR normal). Each column represents an indi- vidual (n=11) and each row a probe set. One hundred and twenty five probe sets were identified using the t-test statistic with an adjusted p value below 0.05 (left heatmap). The 37 probe sets listed on the right represent the specific probe sets which do not reflect the cytogenetic groups, but are specifically associated with risk groups. The variation of absolute expression values is displayed as a varia- tion in color. The color scale extends from –1.13 to 1.11 in log10 space as indicated below the heatmap.
neglect or sexual abuse) (Heim and Nemeroff, 2001; Sadock and Sadock, 2005). In the past years there is also an increasing interest in epigenetic phenomena induced by environmental factors resulting in e.g. histone acetylation/methylation or cytosine methylation. It has been demonstrated in twin studies that these DNA-protein interactions decisively change geneexpression and can be correlated to certain phenotypic changes (Cardno and Gottesman, 2000; Kato et al., 2005). Research on genetic predisposition succeeded in identifying single- nucleotide polymorphisms (SNPs) pointing to a crucial involvement in controlling HPA and SAM axis activity (Ising and Holsboer, 2006). Furthermore, there is an ongoing effort in running unbiased approaches, such as microarray studies or linkage analysis to identify further genes underlying mental disorders. Therefore, it is important to annotate that the genetic blueprint (nature) and the biographic impact (nurture) interact and that in most cases neither one alone can lead to the development of the clinical phenotype (Sillaber and Holsboer, 2004). In some diseases, such as red-green color-vision deficiency, it is well described that the unequal recombination of two pigment genes leads to gene deletion or the formation of hybrid genes that explain the majority of the common red-green color-vision deficiencies (Deeb, 2004). This disease, among others, acts as one example, where a disease or deficiency is restricted to one or a manageable group of genes. Psychiatric diseases, in contrast, are among the most complex diseases due to their multigenic background, with single genes mainly producing small effects and being hard to detect. Pharmacogenomics is not only working on underlying genes and the numerous genetic variations that have been shown to affect disease susceptibility and drug response, but also tries to improve therapy on the basis of genetic information for each patient by focusing on the individual, sex-specific differences, and treatment outcome (Pinsonneault and Sadee, 2003).
Thioredoxin interacting protein (TXNIP) also termed vitamin D3 up-regulated protein 1 (VDUP1) or thioredoxin-binding protein 2 (TBP2) is a 50kDa protein with structural homology to the arrestins . TXNIP originally identified in 1.25-(OH)2D3 stimulated HL60 cells is a stress molecule with multifunctional roles and is implicated in various cellular processes (e.g. oxidative stress, apoptosis, immune function, fatty acid utilization) . It is one of the genes regulated by FOXO1a which regulates the oxidative stress response. TXNIP is able to bind the active domain of thioredoxin (Trx) regulating the function of Trx. TXNIP due to this direct protein-protein interaction inhibits the reducing activity of thioredoxin and therefore the levels of TXNIP influence the vulnerability of the cells to ROS [87, 88]. The over-expression of TXNIP decreases the thioredoxin reducing activity, thus increases the redox stress. Implicated in the development and function of natural killer cells TXNIP shows immunological relevance. Furthermore TXNIP is suggested to deregulate the basic energy metabolism utilizing fatty acid pathways. Tumor cell proliferation and the cell cycle progress are blocked by TXNIP over-expression induced by various growth arrest stimuli. Therefore a strong implication in tumorgenesis and apoptosis is suggested [57, 86].
Genes involved in metabolism were found to be expressed at lower levels in the transferred T cells infiltrating the tumor. PKC γ and DGK α were found to be downregulated in infiltrating T cells. PKC γ is a member of the PKC family that can phosphorylate serine and threonine amino acid residues on the other proteins and control the function of these proteins. Its activation requires increasing concentration of second messengers like diacylglycerol (DAG) and calcium ions (Ca 2+ ) [Mellor et al., 1998]. In general, PKCs are involved in several cell signal transduction cascades, and are upregulated in activated T cells [Isakov et al., 1987, Berry et al., 1990]. Transduction of T cells with PKC γ can up-regulate IL-2 receptor expression, and adoptive transfer of those T cells can lead to tumor regression in vivo [Chen et al., 1994].
After preprocessing, all geneexpressionprofiles were base 10 log-transformed and, in order to prevent single arrays from dominating the analysis, standardized to zero mean and unit variance. In the absence of genuine test sets for four of the six datasets, we performed our benchmark study by repeated ran- dom splitting into learning and test sets exactly as in Dudoit et al. (2002). The data were partitioned into a balanced learn- ing set L comprising two-thirds of the arrays, used for feature preselection, tuning and fitting the classifiers. Then, the class labels of the remaining one-third of the experiments were predicted and the misclassification error was computed as the fraction of predicted class labels that differed from the true one. To reduce the variability, the splitting into learning and test sets was repeated 50 times and the error estimates were averaged. It is important to note that these results are hon- est in the sense that all gene filtering, classifier tuning and fitting operations were re-done on each of the 50 learning sets to allow for reliable conclusions and to avoid overoptimistic results with downward bias.
This mechanism was discovered for the first time in Burkitt’s Lymphoma. To date, three translocations have been associated with this B-cell malignancy. With 90 % of the cases, translocation t(8;14)(q24;q32) is the most frequent. The result is the juxtaposition of cMYC on Chromosome 8 with the immunoglobulin heavy chain (IgH) locus on Chromosome 14 (TH Rabbitts, 1994). The translocation occurs in the switch region of the IgH constant chain gene segments (JL Hecht, 2000; M Nambiar, 2008). In this way, cMYC is relocated near the enhancer of IgH, which leads to an alteration of its expression. cMYC has an important role in cell proliferation, differentiation, apoptosis and metabolism (stem cell renewal). In this way, overexpression of cMYC would have a great impact on multiple cellular processes. This may ultimately contribute to the malignant transformation. Other translocations position cMYC close to immunoglobulin light chain (IgL) κ (t(2;8)(p11;q24)) or IgL λ (t(8;22)(q24;q11)) (TH Rabbitts, 1994, F Mitelman, 2007). These processes are presumably also linked to the switch of recombination of Ig genes. Thus these chromosomal changes do not explain the entire transformation. In addition, mutations in components of the p53 pathway (e.g. p53,
und her bewegt (300 rpm Shaker). Das Bioarray wurde dann gewaschen und mit Streptavidin-Alexa 647 inkubiert. Die Bioarrays wurden mit Hilfe der ScanArray Express Software und einem ScanArray Express HT Scanner (Packard BioScience, Meriden, CT, USA) gescannt, wobei der Laser auf 635 nm, die Laserenergie auf 100% und die Photomultiplier Röhrenspannung auf 60% eingestellt wurden. Die eingescannten Bilddateien wurden mittels CodeLink image and data analysis software (Amersham) untersucht, die sowohl unveränderte als auch normalisierte Hybridisierungs-Signalintensitäten für jeden Arrayspot generierte. Die etwa 10.000 Spotintensitäten des Microarraybildes wurden auf einen Median von 1 standardisiert. Normalisierte Daten mit Signalintensitäten > 0,50 wurden per GeneSifter software (VizX Labs LLC, Seattle, WA, USA, vizxlabs.com) analysiert. Das Programm erstellt auch Reports zur Genontologie und zum z-Score. Diese ontologischen Daten, die entsprechend der Richtlinien des Gene Ontology Consortiums ( http://www.geneontology.org/GO.doc ), 18 organisiert wurden, beziehen sich auf biologische Prozesse, molekulare Funktionen und zelluläre Komponenten.
Cadmium (Cd), one of the main soil pollutants, is a known transcriptional inducer of MTs [ 7 , 11 , 12 ], and therefore, we used it to study transcriptional MT activation. It is known that Cd exerts its toxicity mainly through oxidative stress but also leads to alterations in the DNA methylation pattern [ 14 , 15 ] and epigenetic profiles and phenotypes [ 16 ]. Furthermore, Cd has a negative effect on wound healing mechanisms shown in L. terrestris [ 17 ] and revealed immunosuppressive effects [ 18 – 20 ]. Moreover, Hinrichsen and Tran 2010 found that for Paramecium tetraurelia the sensitivity to Cd has been shown to follow a circadian pattern [ 21 ]. Cahill, Nyberg, and Ehret showed MT levels and Cd accumulation in mice vary during a 24 h period [ 22 ]. Further studies on mice confirmed diurnal MT variations [ 23 ] and its association with Cd sensitivity [ 22 , 24 ]. Our objective was to reveal the MT expression pattern under control conditions, Cd exposure, and physical injury. Moreover, we aimed to elucidate the mechanism of transcriptional MT regulation. The plasticity of MT activation, which was observed upon Cd exposure and physical injury in earthworms, enabled us to determine a region of interest (ROI) in the MT promotor region revealing a selection of transcription factors that might be involved in MT regulation. Additionally, we examined circadian MT expression as well as a putative involvement of promotor DNA methylation.
The study of promoter activities in haloarchaea is carried out exclusively using enzymes as reporters. An alternative reporter is the gene encoding the Green Fluorescent Protein (GFP), a simple and fast tool for investigating promoter strengths. However, the GFP variant smRS-GFP, used to analyze protein stabilities in haloarchaea, is not suitable to quantify weak promoter activities, since the fluorescence signal is too low. We enhanced the fluorescence of smRS-GFP 3.3-fold by introducing ten amino acid substitutions, resulting in mGFP6. Using mGFP6 as reporter, we studied six haloarchaeal promoters exhibiting different promoter strengths. The strongest activity was observed with the housekeeping promoters P fdx of the ferredoxin gene and P2 of the ribosomal 16S rRNA
der Genexpression auch die längste Hypothermiedauer. Die Gewebeprobe unter Hypothermie wurde bei diesem Patienten nach 76 Minuten gewonnen. In dieser Probe wird eine mehr als dop- pelt so hohe UCP1-Expression im Vergleich zum normothermen Gewebe detektiert. Bei Betrach- tung der kompletten Untersuchungsgruppe nimmt er zeitlich jedoch eine Mittelstellung ein. Mit 135 und 120 Minuten am längsten und fast doppelt so lang der Hypothermie ausgesetzt wie Pati- ent Nr. 15, waren die Patienten Nr. 9 und 11. Diese weisen jedoch keinen Anstieg in der Genex- pression, sondern sogar eine Abnahme beziehungsweise keine Änderung der UCP1-Expression auf. Auch bei der Betrachtung drei weiterer Patienten der Gruppe mit UCP1-Induktion ist unter Einbezug des beschriebenen Patienten Nr. 15 kein linearer Zusammenhang zwischen der Hypo- thermiedauer und der Stärke der UCP1-Induktion festzustellen. So weist der Patient (Patient Nr. 13) mit einer vergleichsweise moderaten Induktion um 13 Prozent mit 65 Minuten eine längere Kälteexposition auf, als die Patienten (Patienten Nr. 4, 8), welche bei einer Hypothermiezeit von 58 Minuten eine Zunahme der UCP1-Expression um ein Viertel zeigen. Für den Patienten mit der Nummer 2, welcher eine 30-prozentige Steigerung der Expression von UCP1 aufweist, fehlt die Angabe der Hypothermiedauer. Dieser zeigt als Einziger eine UCP1-Zunahme bei schon relativ ho- her basaler UCP1-Expression von relativ zur TAF1-Expression errechneten 738 Kopien. Die drei weiteren Patienten mit hohen basalen UCP1-Expressionsraten zeigen in Gewebeprobe Ⅱ eine Ab- nahme der Genexpression. Verschiedene Studien belegen, dass Kälte ein Aktivator der Thermoge- nese in braunen und beigen Adipozyten ist (van der Lans et al., 2013, Saito et al., 2009a, Vitali et al., 2012a). β-adrenerg stimuliert werden vermehrt Proteine synthetisiert, welche für die Wärme- bildung unerlässlich sind. Zu diesen gehört UCP1. Die mRNA als Transkriptionsprodukt ließe sich vermehrt messen. Folglich ist eine detaillierte Analyse der Gruppe, welche eine Abnahme der
integrating environmental and genetic cues have not yet been fully elucidated. We aimed to test, whether the allele of an intronic 3.3 kb large variant esv3608688 (insertion / deletion), located within in a disease-associated haplotype at the FKBP5 locus, have an effect on glucocorticoid-dependent gene regulation and how this can be explained on a molecular level. We observed that the deletion allele was associated with a higher induction of FKBP5 expression after activation of the glucocorticoid receptor (GR) in lymphoblastoid cell lines (LCLs). Furthermore, we could show that the esv3608688 allele can moderate the effect on FKBP5 geneexpression regulation of a previously described functional single nucleotide polymorphism rs1360780 (Klengel et al, 2013). We identified that the stabilization of architectural and enhancer-promoter loops is a common feature of the factors (T- allele, deletion and GR activation) leading to increased FKBP5 mRNA expression. In addition, we observed differentially methylated sites depending on the allele status for esv3608688 within key regulatory sites of FKBP5. Our data proposes a model, that this geneexpression response is most likely result of an increased activity of GRE enhancers due to the stabilization of chromatin interactions. This study highlights molecular mechanism how genetic and environmental factors can be integrated fine-tune FKBP5 expression levels, potentially involved in leading to FKBP5 disinhibition and therefore shape the risk for developing a stress-related psychiatric disease.
O’C ONNOR et al. (2003) zeigten in ihrer Studie, dass die CIAS1/NALP3-Expression in humanen Monozyten durch eine Vielzahl pro-inflammatorischer Stimuli wie TNF-α und TLR-Liganden induziert wird. CIAS1 ist in Abwesenheit von ASC in der Lage, das Signalgeschehen von NF-κB negativ zu regulieren und somit das Ausmaß entzündlichen Reaktionen zu beschränken (O’C ONNOR et al., 2003). Zusammenfassend kann CIAS1/NALP3 in Verbindung mit ASC ein Inflammasom bilden und somit auto-inflammatorische Prozesse verstärken, während es ohne die Assoziation mit PYCARD/ASC diese Prozesse vermindern kann. Eine starke Herunterregulation von NALP3 im hohen Alter könnte dazu führen, dass inflammatorische Prozesse sowie Zytokinbildung vermindert initiiert werden und damit die physiologische Abwehrreaktion verringert wird.
Background: Lactobacillus plantarum constitutes a well-recognized food-grade system for the expression of recombinant proteins in the field of industrial and medical biotechnology. For applications in vivo or in biotechnological processes, the level of expression of e.g. antigens or enzymes is often critical, as expression levels should be of a certain effectiveness, yet, without putting too much strain to the overall system. The key factors that control geneexpression are promoter strength, gene copy number and translation efficiency. In order to estimate the impact of these adjusting screws in L. plantarum CD033, we have tested several constitutive promoters in combination with high and low copy number plasmid backbones and varying space between the Shine-Dalgarno sequence and the start-codon.
Several specific miRNAs have already been associated with WTs [29,33–35]. Senanayake et al.  compared the expression of five miRNAs (miR-192, -194, -215, -200c, and -141), each between a normal kidney and WT of different histological subtypes. They found a significant downregulation of miR-192, miR-215 and miR-194 in all Wilms tumors irrespective of the subtype, and further a significant downregulation of miR-141 and miR-200c in blastemal and mixed WTs compared to normal kidneys . Our findings support this result, as all of these miRNAs were significantly downregulated 21- to 46-fold in our WT set compared to a normal kidney. Kort et al.  demonstrated a significant overexpression of the oncomiR-1 cluster containing miR-17, -18a, -19b, -20a and -92 in WTs compared to normal kidneys, and an overexpression of several miRNAs, including miR-130b and miR-181b, in WTs compared to other renal tumors . In our data, we found an overexpression of miR-18a and miR-130b in blastemal WTs and miR-181b in regressive WTs, each compared to normal kidneys. Furthermore, we could confirm the overexpression of miR-483-3p in WTs compared to normal kidneys. Previously, this overexpression was also shown by Veronese et al. . There is also a considerable overlap of deregulated miRNAs in our study with the study of Watson et al. . The results of Watson et al.  are, however, difficult to compare to our study. Watson et al.  compared miRNA expression in the blastemal component of high-risk patients with either the blastemal or the non-blastemal component of intermediate risk patients, and did not include normal kidney tissue as a reference. By contrast, our study focused on the differential expression of miRNA in blastemal (high-risk) and regressive (intermediate-risk) subtypes in comparison to normal kidneys, and without further information on the blastemal cell content of our probes. Of the 14 miRNAs reported to be upregulated in the high-risk blastemal compared to intermediate-risk non-blastemal component, we found that three miRNAs, namely miR-590-5p, miR-125a-5p and miR-19a, upregulated in blastemal WTs as compared to the regressive subtype. This analysis comes closest to the comparison of Watson et al. . By contrast, out of the 17 reported miRNAs that were downregulated in high-risk blastemas, eight were also identified as deregulated in our study, with six miRNAs showing the opposite direction of deregulation, i.e., an increased expression in blastemal WTs. Without data on the exact cellular composition of the regressive tumor samples used in our study, this obvious discrepancy could be attributed to possible residual blastemal cells in our regressive tumors.
We propose modifications and extensions of boosting clas- sifiers for microarray geneexpression data from several tissue or cancer types. We applied precedent feature selec- tion and used the more robust LogitBoost combined with an alternative approach for binary problems. The results on six real and a simulated datasets indicate that these modifications are successful and make boosting a com- petitive player for predicting expression data. Our feature preselection generally improved the predictive power of a classifier. Moreover, we observed slightly better perfor- mance of LogitBoost over AdaBoost, and our whole pro- cedure (feature selection plus LogitBoost) compares fa- vorably with previously published results using AdaBoost. Finally, we propose to reduce multiclass problems to mul- tiple binary problems which are solved separately. This was found to have a great potential for more accurate re- sults on geneexpression data, where the choice of predic- tor variables is crucial.
Reads were mapped onto the human reference genome release hg38 (GRCh38) [ 72 ] with Ensembl transcript annotation version 87 [ 48 ] using Tophat version 2.1.1 [ 73 ] with Bowtie version 2.2.9 [ 74 ]. Reads were counted with featureCounts [ 75 ] and geneexpression values (reads per kilobase exon per million mapped reads (RPKM)) were calculated with Cufflinks version 2.2 [ 76 ]. The differential expression between two sample groups was calculated with edgeR [ 77 ]. The filtering for differentially expressed genes is for p-value of 0.05 (FWER cor- rected) and minimal fold-change of 2. In the more specific analyses for single genes, for the dif- ferences in men and woman and for classification models, (healthy) age-related genes are removed. This is because the sampled healthy subjects are in average quite younger than the subjects with different arthritis conditions and at comparisons between them age-related genes are expected to be significantly different. Age-related genes are taken from Yang et. al. [ 78 ]. We performed also comparisons of geneexpression between groups adjusted for age. At the most changing adjustment in the comparison between healthy subjects and early RA (an average age of 35.2 vs. 55.9), we realized that many genes well known for RA are filtered (as CCL19 [ 79 ], CCL22 [ 80 ], CCR6 [ 81 ], CD6 [ 82 ], CDH11 [ 83 ], IFIT1B (as a paralog to IFIT1 [ 84 ]), IL26 [ 85 ], IL2RB [ 86 ], MMP10 [ 87 ], MMP12 [ 88 ], MMP8 [ 89 ] and MMP9 [ 90 ]). Similar worrying are the overlaps between unique DEGs in the comparison unadjusted and adjusted by age with the external age-related genes (as used for filtering from Yang et al. [ 78 ]), we see even a higher overlap between age-adjusted DEGs (healthy vs early RA) and the external age- related genes. Given that, we used the comparison without adjustment for further analyses. Age-adjusted comparisons are available in the Supplementary Archive.
All MCC tissues and cell lines in this study were negative for ASCL1 expression except one MCC which revealed a focal expression of ASCL1 (Table 1). Ralston and colleagues (2008) have previously reported that MCCs were all tested negative for ASCL1 expression by IHC . In contrast, Lewis et al. (2010) found that MCCs arising in the head and neck region were positive for ASCL1 expression . Interestingly, the anti-ASCL antibody in this study and the two previous reports derived from the same antibody clone. Moreover, we show that ASCL1 expression is also absent in all tested MCC cell lines, irrespective of the MCPyV-status (Figure 2).
Advanced methods use libraries of constructs, transcribed and translated with different strengths by varying the promoter ( Kraft et al., 2007 ), the ribosome binding site ( Kohl et al., 2018 ) or by fusion with peptides aiding in correct folding ( Butt et al., 2005; Kraft et al., 2007 ). Another approach is to vary the copy number of the genetic constructs. As shown before, the number in which a gene is present in a cell is correlated with the expression level of said gene ( Lee et al., 2015 ). In the most prominent host Escherichia coli, plasmids with different copy numbers are available such as the high copy number pUC vectors ( Yanisch-Perron et al., 1985 ) and the low copy number pBR322 based vectors ( Bolivar et al., 1977 ) and assisted integration into the chromosome is possible through recombineering ( Sharan et al., 2009 ). Plasmid replication is a major burden on the cell’s metabolism, thus always requiring an active selection mechanism to ensure plasmid stability during production. The higher the copy number of the plasmids, the higher this burden and the less stable the plasmid DNA is propagated. For the yeast S. cerevisiae it was demonstrated, that a set of tunable copy number plasmids (dependent on antibiotic concentration) could balance a pathway for n-butanol production in a way that yielded a 100-fold increase in production of the desired chemical ( Lian et al., 2016 ). In another publication, different copy number plasmids were introduced into Bacillus methanolicus via electroporation to show the effect of gene dosage on the expression of gfpuv, an amylase from Streptomyces griseus and the lysine decarboxylase cadA from E. coli for overproduction of cadaverine. A positive correlation between estimated plasmid copy numbers and observed expression levels could be determined ( Irla et al., 2016 ). Since plasmid copy numbers can be affected by media composition as well as growth phase, efforts have been undertaken to uncouple expression levels from
s norm, Euclidean distance, Manhattan distance and dynamic time warping with the step pattern symmetric1, symmetric2 and asymmetric. Figure 8: The role of interpolation and sampling: simulated expression time series of 100 equally sampled data points (black line), the effect of (spline) interpolation (including the following data points of the original series: 1|2|3|6|9|15|25|39|63|99., green line). Figure 9: Artefacts introduced in the reconstruction procedure (measure: μ I , scoring scheme: ID) by interpolation of short, coarsely sampled time series. The left panel shows the corresponding ROC curves in the noise-free case for 10 points equally sampled in time, whereas the right panel presents the same results for 10 points, unequally sampled. The unequal sampling in time is the same as in Figure 8. Figure 10: ROC curves for selected measures and algorithms obtained in the noise-free case, using unequally sampled data without interpolation. The sampling is the same as in the previous two figures, including the following data points of a simulated series of 100 points: 1|2|3|6|9|15|25|39|63|99. Figure 11: ROC curves obtained from the reconstruction of an E. coli network of 100 genes, a S.cerevisiae network of 100 gene and an E. coli network of 200 genes. (a)-(i) show the results using various similarity measures together with the ID scoring scheme: (a) Euclidean distance μ EC , (b) Manhattan distance μ MA , (c) Ls norm μ L , (d) Kendall ’s rank correlation μ K , (e) Pearson correlation μP, (f) conditional Pearson correlation μ c P , (g) mutual information of symbol vectors μ I