The g3mclass is a practical software for multiclass classification on biomarkers.

The analytes qualified as biomarkers are potent tools to diagnose various diseases, monitor therapy responses, and design therapeutic interventions. The early assessment of the diverseness of human disease is essential for the speedy and cost-efficient implementation of personalized medicine. We developed g3mclass, the Gaussian mixture modeling software for molecular assay data classification. This software automates the validated multiclass classifier applicable to single analyte tests and multiplexing assays. The g3mclass achieves automation using the original semi-constrained expectation-maximization (EM) algorithm that allows inference from the test, control, and query data that human experts cannot interpret. In this study, we used real-world clinical data and gene expression datasets (ERBB2, ESR1, PGR) to provide examples of how g3mclass may help overcome the problems of over-/underdiagnosis and equivocal results in diagnostic tests for breast cancer. We showed the g3mclass output's accuracy, robustness, scalability, and interpretability. The user-friendly interface and free dissemination of this multi-platform software aim to ease its use by research laboratories, biomedical pharma, companion diagnostic developers, and healthcare regulators. Furthermore, the g3mclass automatic extracting information through probabilistic modeling is adaptable for blending with machine learning and artificial intelligence.

IsoSolve: An Integrative Framework to Improve Isotopic Coverage and Consolidate Isotopic Measurements by Mass Spectrometry and/or Nuclear Magnetic Resonance.

Stable-isotope labeling experiments are widely used to investigate the topology and functioning of metabolic networks. Label incorporation into metabolites can be quantified using a broad range of mass spectrometry (MS) and nuclear magnetic resonance (NMR) spectroscopy methods, but in general, no single approach can completely cover isotopic space, even for small metabolites. The number of quantifiable isotopic species could be increased and the coverage of isotopic space improved by integrating measurements obtained by different methods; however, this approach has remained largely unexplored because no framework able to deal with partial, heterogeneous isotopic measurements has yet been developed. Here, we present a generic computational framework based on symbolic calculus that can integrate any isotopic data set by connecting measurements to the chemical structure of the molecules. As a test case, we apply this framework to isotopic analyses of amino acids, which are ubiquitous to life, central to many biological questions, and can be analyzed by a broad range of MS and NMR methods. We demonstrate how this integrative framework helps to (i) clarify and improve the coverage of isotopic space, (ii) evaluate the complementarity and redundancy of different techniques, (iii) consolidate isotopic data sets, (iv) design experiments, and (v) guide future analytical developments. This framework, which can be applied to any labeled element, isotopic tracer, metabolite, and analytical platform, has been implemented in IsoSolve (available at https://github.com/MetaSys-LISBP/IsoSolve and https://pypi.org/project/IsoSolve), an open-source software that can be readily integrated into data analysis pipelines.

Exploring the Glucose Fluxotype of the E. coli y-ome Using High-Resolution Fluxomics.

We have developed a robust workflow to measure high-resolution fluxotypes (metabolic flux phenotypes) for large strain libraries under fully controlled growth conditions. This was achieved by optimizing and automating the whole high-throughput fluxomics process and integrating all relevant software tools. This workflow allowed us to obtain highly detailed maps of carbon fluxes in the central carbon metabolism in a fully automated manner. It was applied to investigate the glucose fluxotypes of 180 Escherichia coli strains deleted for y-genes. Since the products of these y-genes potentially play a role in a variety of metabolic processes, the experiments were designed to be agnostic as to their potential metabolic impact. The obtained data highlight the robustness of E. coli's central metabolism to y-gene deletion. For two y-genes, deletion resulted in significant changes in carbon and energy fluxes, demonstrating the involvement of the corresponding y-gene products in metabolic function or regulation. This work also introduces novel metrics to measure the actual scope and quality of high-throughput fluxomics investigations.

Theoretical Basis for Dynamic Label Propagation in Stationary Metabolic Networks under Step and Periodic Inputs.

The dynamics of label propagation in a stationary metabolic network during an isotope labeling experiment can provide highly valuable information on the network topology, metabolic fluxes, and on the size of metabolite pools. However, major issues, both in the experimental set-up and in the accompanying numerical methods currently limit the application of this approach. Here, we propose a method to apply novel types of label inputs, sinusoidal or more generally periodic label inputs, to address both the practical and numerical challenges of dynamic labeling experiments. By considering a simple metabolic system, i.e. a linear, non-reversible pathway of arbitrary length, we develop mathematical descriptions of label propagation for both classical and novel label inputs. Theoretical developments and computer simulations show that the application of rectangular periodic pulses has both numerical and practical advantages over other approaches. We applied the strategy to estimate fluxes in a simulated experiment performed on a complex metabolic network (the central carbon metabolism of Escherichia coli), to further demonstrate its value in conditions which are close to those in real experiments. This study provides a theoretical basis for the rational interpretation of label propagation curves in real experiments, and will help identify the strengths, pitfalls and limitations of such experiments. The cases described here can also be used as test cases for more general numerical methods aimed at identifying network topology, analyzing metabolic fluxes or measuring concentrations of metabolites.

Correction of MS data for naturally occurring isotopes in isotope labelling experiments.

Mass spectrometry (MS) in combination with isotope labelling experiments is widely used for investigations of metabolism and other biological processes. Quantitative applications-e.g., (13)C metabolic flux analysis-require correction of raw MS data (isotopic clusters) for the contribution of all naturally abundant isotopes. This chapter describes how to perform such correction using the software IsoCor. This flexible, user-friendly software can be used to exploit any isotopic tracer, from well-known ((13)C, (15)N, (18)O, etc.) to unusual ((57)Fe, (77)Se, etc.) isotopes. It also provides options-e.g., correction for the isotopic purity of the tracer-to improve the accuracy of quantitative isotopic studies, and allows automated correction of large datasets that can be collected with modern MS methods.

A novel platform for automated high-throughput fluxome profiling of metabolic variants.

Advances in metabolic engineering are enabling the creation of a large number of cell factories. However, high-throughput platforms do not yet exist for rapidly analyzing the metabolic network of the engineered cells. To fill the gap, we developed an integrated solution for fluxome profiling of large sets of biological systems and conditions. This platform combines a robotic system for (13)C-labelling experiments and sampling of labelled material with NMR-based isotopic fingerprinting and automated data interpretation. As a proof-of-concept, this workflow was applied to discriminate between Escherichia coli mutants with gradual expression of the glucose-6-phosphate dehydrogenase. Metabolic variants were clearly discriminated while pathways that support metabolic flexibility towards modulation of a single enzyme were elucidating. By directly connecting the data flow between cell cultivation and flux quantification, considerable advances in throughput, robustness, release of resources and screening capacity were achieved. This will undoubtedly facilitate the development of efficient cell factories.

IsoDesign: a software for optimizing the design of 13C-metabolic flux analysis experiments.

The growing demand for (13) C-metabolic flux analysis ((13) C-MFA) in the field of metabolic engineering and systems biology is driving the need to rationalize expensive and time-consuming (13) C-labeling experiments. Experimental design is a key step in improving both the number of fluxes that can be calculated from a set of isotopic data and the precision of flux values. We present IsoDesign, a software that enables these parameters to be maximized by optimizing the isotopic composition of the label input. It can be applied to (13) C-MFA investigations using a broad panel of analytical tools (MS, MS/MS, (1) H NMR, (13) C NMR, etc.) individually or in combination. It includes a visualization module to intuitively select the optimal label input depending on the biological question to be addressed. Applications of IsoDesign are described, with an example of the entire (13) C-MFA workflow from the experimental design to the flux map including important practical considerations. IsoDesign makes the experimental design of (13) C-MFA experiments more accessible to a wider biological community. IsoDesign is distributed under an open source license at http://metasys.insa-toulouse.fr/software/isodes/

IsoCor: correcting MS data in isotope labeling experiments.

UNLABELLED: Mass spectrometry (MS) is widely used for isotopic labeling studies of metabolism and other biological processes. Quantitative applications-e.g. metabolic flux analysis-require tools to correct the raw MS data for the contribution of all naturally abundant isotopes. IsoCor is a software that allows such correction to be applied to any chemical species. Hence it can be used to exploit any isotopic tracer, from well-known ((13)C, (15)N, (18)O, etc) to unusual ((57)Fe, (77)Se, etc) isotopes. It also provides new features-e.g. correction for the isotopic purity of the tracer-to improve the accuracy of quantitative isotopic studies, and implements an efficient algorithm to process large datasets. Its user-friendly interface makes isotope labeling experiments more accessible to a wider biological community. AVAILABILITY: IsoCor is distributed under OpenSource license at http://metasys.insa-toulouse.fr/software/isocor/

influx_s: increasing numerical stability and precision for metabolic flux analysis in isotope labelling experiments.

MOTIVATION: The problem of stationary metabolic flux analysis based on isotope labelling experiments first appeared in the early 1950s and was basically solved in early 2000s. Several algorithms and software packages are available for this problem. However, the generic stochastic algorithms (simulated annealing or evolution algorithms) currently used in these software require a lot of time to achieve acceptable precision. For deterministic algorithms, a common drawback is the lack of convergence stability for ill-conditioned systems or when started from a random point. RESULTS: In this article, we present a new deterministic algorithm with significantly increased numerical stability and accuracy of flux estimation compared with commonly used algorithms. It requires relatively short CPU time (from several seconds to several minutes with a standard PC architecture) to estimate fluxes in the central carbon metabolism network of Escherichia coli. AVAILABILITY: The software package influx_s implementing this algorithm is distributed under an OpenSource licence at http://metasys.insa-toulouse.fr/software/influx/. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

An approach to the study of gene expression in hepatocarcinogenesis initiation.

In carcinogenesis, determination of gene and protein expression profiles is important for prevention and treatment. Caffeic acid phenethyl ester (CAPE) in a single dose administered before carcinogenic initiation induced by diethylnitrosamine (DEN) prevents the appearance of preneoplastic lesions. On the basis of this approach, the main purpose of this work was to compare the gene expression profiles induced by DEN or a previously administered single dose of CAPE. Using a modified hepatocarcinogenesis-resistant hepatocyte model, male Fischer-344 rats were administered with one intraperitoneal dose of CAPE (20 mg/kg) 12 hours before DEN administration (200 mg/kg). Livers were removed and processed for microarray analysis and reverse transcription-polymerase chain reaction 12 hours after CAPE dosing and 24 hours after DEN administration with or without CAPE. CAPE alone did not alter the expression profile. DEN treatment modified the expression of 665 genes, and CAPE plus DEN induced changes in 1371 genes. DEN treatment increased the expression of genes associated with oxidative stress such as glutathione reductase, genes involved in cell cycle regulation including p53, and modified cytochrome P450. CAPE plus DEN diminished the expression of cytochrome involved in DEN bioactivation such as CYP2B1 as well as the expression of regulators of oxidative stress such as glutathione reductase, GST-kappa and GST-theta, and cell cycle regulators such as p53. Using CAPE as a tool, we uncovered new approaches for studying the altered expression of reactive genes and identifying proteins that will help to propose well-sustained and concrete hypothesis of DEN mechanism of hepatocarcinogenesis initiation.

Celecoxib activates Stat5 and restores or increases the expression of growth hormone-regulated genes in hepatocarcinogenesis.

We have previously evaluated the chemopreventive effect of celecoxib on preneoplastic lesions in rat liver. However, though the effects of celecoxib have been tested in a variety of carcinomas, there has not been a study on the modulation of gene expression in response to this drug. Here, we evaluated the effect of celecoxib on the gene expression profile associated with hepatocarcinogenesis. Male Sprague-Dawley rats underwent the modified resistant hepatocyte model and were fed a diet containing 1500 ppm of celecoxib. Gene expression profiles were evaluated using DNA microarrays and further validations were performed using quantitative PCR, western blotting and immunohistochemical staining. Celecoxib modulated the expression of 46 genes, and those regulated by growth hormone were selected for further analysis. Celecoxib significantly upregulated the expression of the Cyp2b1/2, Cyp3a1, and alpha2-urinary globulin (alpha2uG) genes and restored the expression of Cyp2b3 to normal. The protein expression of Cyp2b1/2 was increased, but the expressions of Cyp3a1 and alpha2uG were only restored to normal levels. The increased Cyp2b1/2 expression in response to celecoxib was mainly confined to preneoplastic lesions. A search for the upstream mediator of these genetic alterations found that carcinogenesis inactivated by 87% the signal transducer and activator of transcription 5 (Stat5), a transcription factor that is activated by growth hormone signaling, but celecoxib treatment restored its activation. In conclusion, these results suggest that celecoxib exerts anticancer effects on altered hepatic cells by restoring mRNA and the protein expression levels of specific genes, in part through the reactivation of Stat5.

The YTA7 gene is involved in the regulation of the isoprenoid pathway in the yeast Saccharomyces cerevisiae.

The isoprenoid pathway in yeasts is important not only for sterol biosynthesis but also for the production of nonsterol molecules, deriving from farnesyl diphosphate (FPP), implicated in N-glycosylation and biosynthesis of heme and ubiquinones. FPP formed from mevalonate in a reaction catalyzed by FPP synthase (Erg20p). In order to investigate the regulation of Erg20p in Saccharomyces cerevisiae, we searched for its protein partners using a two-hybrid screen, and identified five interacting proteins, among them Yta7p. Subsequently, we showed that Yta7p was a membrane-associated protein localized both to the nucleus and to the endoplasmic reticulum. Deletion of YTA7 affected the enzymatic activity of cis-prenyltransferase (the enzyme that utilizes FPP for dolichol biosynthesis) and the cellular levels of isoprenoid compounds. Additionally, it rendered cells hypersensitive to lovastatin, an inhibitor of 3-hydroxy-3-methylglutaryl coenzyme A reductase (HMGR) that acts upstream of FPP synthase in the isoprenoid pathway. While HMGR is encoded by two genes, HMG1 and HMG2, only HMG2 overexpression was able to restore growth of the yta7Delta cells in the presence of lovastatin. Moreover, the expression level of the S. cerevisiae YTA7 gene was altered upon impairment of the isoprenoid pathway not only by lovastatin but also by zaragozic acid, an inhibitor of squalene synthase. Altogether, these results provide substantial evidence of Yta7p involvement in the regulation of isoprenoid biosynthesis.

Comparative transcriptome analysis between original and evolved recombinant lactose-consuming Saccharomyces cerevisiae strains.

The engineering of Saccharomyces cerevisiae strains for lactose utilization has been attempted with the intent of developing high productivity processes for alcoholic fermentation of cheese whey. A recombinant S. cerevisiae flocculent strain that efficiently ferments lactose to ethanol was previously obtained by evolutionary engineering of an original recombinant that displayed poor lactose fermentation performance. We compared the transcriptomes of the original and the evolved recombinant strains growing in lactose, using cDNA microarrays. Microarray data revealed 173 genes whose expression levels differed more than 1.5-fold. About half of these genes were related to RNA-mediated transposition. We also found genes involved in DNA repair and recombination mechanisms, response to stress, chromatin remodeling, cell cycle control, mitosis regulation, glycolysis and alcoholic fermentation. These transcriptomic data are in agreement with some of the previously identified physiological and molecular differences between the recombinants, and point to further hypotheses to explain those differences.

Transcriptomic analysis of the exit from dormancy of Aspergillus fumigatus conidia.

BACKGROUND: Establishment of aspergillosis is depending upon the exit from dormancy and germination of the conidia of Aspergillus fumigatus in the lung. To gain an understanding of the molecular mechanisms underlying the early steps of conidial germination, we undertook a transcriptomic analysis using macroarrays constructed with PCR fragments from > 3,000 genes (around one third of the annotated A. fumigatus genome). RESULTS: Major results of this analysis are the following: (i) conidia stored pre-packaged mRNAs transcripts (27% of genes have transcripts in the resting conidia; (ii) incubation at 37 degrees C in a nutritive medium induced up- and down-regulation of genes: 19% of the total number of genes deposited on the array were up-regulated whereas 22% of the genes with pre-packaged mRNA in the resting conidia were down-regulated; (iii) most modifications were seen during the first 30 min of germination whereas very little modification of gene expression occurred during the following hour; (iv) one-year old conidia and one-week old conidia behaved similarly at transcriptional level. CONCLUSION: Transcriptomic data indicate that the exit from dormancy is associated with a shift from a fermentative metabolism to a respiratory metabolism as well as a trend toward immediate protein synthesis.

Early transcriptional response of wine yeast after rehydration: osmotic shock and metabolic activation.

The inoculation of active dry wine yeast (ADWY) is one of the most common practices in winemaking. We have used DNA microarray technology to examine the genetic expression patterns of a commercial ADWY strain after rehydration. After rehydration of ADWY for 30 min, a further hour in water after rehydration did not lead to any relevant changes in global gene expression. Expression changes in rehydrated cells upon incubation in a sorbitol solution at the same osmotic pressure as in complete must were rather limited, whereas the presence of fermentable carbon sources or the complete medium (synthetic must) produced very similar transcriptional responses. The main responses were the activation of some genes of the fermentation pathway and of the nonoxidative branch of the pentose phosphate pathway, and the induction of a huge cluster of genes related to ribosomal biogenesis and protein synthesis. The presence of cycloheximide in fermentable medium produced a similar but stronger transcriptional response. Whereas the viabilities of rehydrated cells incubated for 1 h in these different media were similar, yeast vitality, which represents the fermentative capacity of the yeast, showed a positive correlation with the availability of a fermentable carbon source.

Integration of transcriptomic and metabolic analyses for understanding the global responses of low-temperature winemaking fermentations.

Wine produced at low temperature is often considered to have improved sensory qualities. To investigate the effects of temperature on winemaking, the expression patterns during the industrial fermentation process carried out at 13 degrees C and 25 degrees C were compared, and correlated with physiological and biochemical data, including viability, fermentation byproducts and lipid content of the cells. From a total of 535 ORFs that were significantly differentially expressed between the 13 degrees C and 25 degrees C fermentations, two significant transcription programmes were identified. A cold-stress response was expressed at the initial stage of the fermentation, and this was followed by a transcription pattern of upregulated genes concerned with the cell cycle, growth control and maintenance in the middle and late stages of the process at 13 degrees C with respect to 25 degrees C. These expression patterns were correlated with higher cell viability at low temperature. The other relevant transcriptomic difference was that several genes implicated in cytosolic fatty acid synthesis were downregulated, while those involved in mitochondrial short-chain fatty acid synthesis were upregulated in the fermentation process conducted at 13 degrees C with respect to that at 25 degrees C. These transcriptional changes were qualitatively correlated with improved resistance to ethanol and increased production of short-chain (C(4)-C(8)) fatty acids and their corresponding esters at 13 degrees C as compared to 25 degrees C. While this increase of ethyl esters may account in part for the improved sensory quality of wine fermented at 13 degrees C, it is still unclear how the esterification of the short-chain fatty acids takes place. On the basis of its strong upregulation at 13 degrees C, we propose a possible role of IAH1 encoding an esterase/ester synthase in this process.

Investigating the caffeine effects in the yeast Saccharomyces cerevisiae brings new insights into the connection between TOR, PKC and Ras/cAMP signalling pathways.

Caffeine is a natural purine analogue that elicits pleiotropic effects leading ultimately to cell's death by a largely uncharacterized mechanism. Previous works have shown that this drug induces a rapid phosphorylation of the Mpk1p, the final mitogen-activated protein (MAP) kinase of the Pkc1p-mediated cell integrity pathway. In this work, we showed that this phosphorylation did not necessitate the main cell wall sensors Wsc1p and Mid2p, but was abolished upon deletion of ROM2 encoding a GDP/GTP exchange factor of Rho1p. We also showed that the caffeine-induced phosphorylation of Mpk1p was accompanied by a negligible activation of its main downstream target, the Rlm1p transcription factor. This result was consolidated by the finding that the loss of RLM1 had no consequence on the increased resistance of caffeine-treated cells to zymolyase, indicating that the cell wall modification caused by this drug is largely independent of transcriptional activation of Rlm1p-regulated genes. Additionally, the transcriptional programme elicited by caffeine resembled that of rapamycin, a potent inhibitor of the TOR1/2 kinases. Consistent with this analysis, we found that the caffeine-induced phosphorylation of Mpk1p was lost in a tor1Delta mutant. Moreover, a tor1Delta mutant was, like mutants defective in components of the Pkc1p-Mpk1p cascade, highly sensitive to caffeine. However, the hypersensitivity of a tor1 null mutant to this drug was rescued neither by sorbitol nor by adenine, which was found to outcompete caffeine effects specially on mutants in the PKC pathway. Altogether, these data indicated that Tor1 kinase is a target of caffeine, whose inhibition incidentally activates the Pkc1p-Mpk1p cascade, and that the caffeine-dependent phenotypes are largely dependent on inhibition of Tor1p-regulated cellular functions. Finally, we found that caffeine provoked, in a Rom2p-dependent manner, a transient drop in intracellular levels of cAMP, that was followed by change in expression of genes implicated in Ras/cAMP pathway. This result may pose Rom2p as a mediator in the interplay between Tor1p and the Ras/cAMP pathway.

Gene expression profile related to the progression of preneoplastic nodules toward hepatocellular carcinoma in rats.

In this study, we investigated the time course gene expression profile of preneoplastic nodules and hepatocellular carcinomas (HCC) to define the genes implicated in cancer progression in a resistant hepatocyte model. Tissues that included early nodules (1 month, ENT-1), persistent nodules (5 months, ENT-5), dissected HCC (12 months), and normal livers (NL) from adult rats were analyzed by cDNA arrays including 1185 rat genes. Differential genes were derived in each type of sample (n = 3) by statistical analysis. The relationship between samples was described in a Venn diagram for 290 genes. From these, 72 genes were shared between tissues with nodules and HCC. In addition, 35 genes with statistical significance only in HCC and with extreme ratios were identified. Differential expression of 11 genes was confirmed by comparative reverse transcription-polymerase chain reaction, whereas that of 2 genes was confirmed by immunohistochemistry. Members involved in cytochrome P450 and second-phase metabolism were downregulated, whereas genes involved in glutathione metabolism were upregulated, implicating a possible role of glutathione and oxidative regulation. We provide a gene expression profile related to the progression of nodules into HCC, which contributes to the understanding of liver cancer development and offers the prospect for chemoprevention strategies or early treatment of HCC.

Dendrimeric coating of glass slides for sensitive DNA microarrays analysis.

Successful use and reliability of microarray technology is highly dependent on several factors, including surface chemistry parameters and accessibility of cDNA targets to the DNA probes fixed onto the surface. Here, we show that functionalisation of glass slides with homemade dendrimers allow production of more sensitive and reliable DNA microarrays. The dendrimers are nanometric structures of size-controlled diameter with aldehyde function at their periphery. Covalent attachment of these spherical reactive chemical structures on amino-silanised glass slides generates a reactive approximately 100 A layer onto which amino-modified DNA probes are covalently bound. This new grafting chemistry leads to the formation of uniform and homogenous spots. More over, probe concentration before spotting could be reduced from 0.2 to 0.02 mg/ml with PCR products and from 20 to 5 micro M with 70mer oligonucleotides without affecting signal intensities after hybridisation with Cy3- and Cy5-labelled targets. More interestingly, while the binding capacity of captured probes on dendrimer-activated glass surface (named dendrislides) is roughly similar to other functionalised glass slides from commercial sources, detection sensitivity was 2-fold higher than with other available DNA microarrays. This detection limit was estimated to 0.1 pM of cDNA targets. Altogether, these features make dendrimer-activated slides ideal for manufacturing cost-effective DNA arrays applicable for gene expression and detection of mutations.