Web-based analysis of the mouse transcriptome using Genevestigator
© Laule et al; licensee BioMed Central Ltd. 2006
Received: 31 March 2006
Accepted: 21 June 2006
Published: 21 June 2006
Gene function analysis often requires a complex and laborious sequence of laboratory and computer-based experiments. Choosing an effective experimental design generally results from hypotheses derived from prior knowledge or experimentation. Knowledge obtained from meta-analyzing compendia of expression data with annotation libraries can provide significant clues in understanding gene and network function, resulting in better hypotheses that can be tested in the laboratory.
Genevestigator is a microarray database and analysis system allowing context-driven queries. Simple but powerful tools allow biologists with little computational background to retrieve information about when, where and how genes are expressed. We manually curated and quality-controlled 3110 mouse Affymetrix arrays from public repositories. Data queries can be run against an annotation library comprising 160 anatomy categories, 12 developmental stage groups, 80 stimuli, and 182 genetic backgrounds or modifications. The quality of results obtained through Genevestigator is illustrated by a number of biological scenarios that are substantiated by other types of experimentation in the literature.
The Genevestigator-Mouse database effectively provides biologically meaningful results and can be accessed at https://www.genevestigator.ethz.ch.
The development of functional genomics technologies has led in recent years to a proliferation of databases for storage and delivery of microarray data. Since the introduction of the MIAME standard  and associated community- level annotation guidelines [2–4], experimental descriptions have become more precise, allowing a better understanding and reproducibility of experiments, as well as more efficient querying possibilities. Several databases offer tools to browse, query and download experiments. However, in most cases, data is provided as is, without removal of biased data after systematic processing with quality-control measures. Furthermore, the focus of most microarray databases has been in storage and retrieval of experiments, but only few provide analysis tools optimally interacting with the database. In parallel to these developments, several web-based tools have recently been developed specifically for the analysis of individual microarray experiments, such as RACE  or ArrayQuest . High-throughput technologies allow to streamline the same type of analysis for large numbers of genes or proteins. A major challenge for scientists in this respect is the sparsity of the data sets, i.e. the low number of measurements relative to the immense number of simultaneously tested elements. The analysis of such data structures often cannot make use of many classical statistical procedures and calls for the development of novel statistical approaches, such as sparse graphical modeling  or computational approaches that allow to compile result summaries combining data and annotations.
Genevestigator  is a high-quality database combined with tools to create such result summaries. It reveals novel and diverse information about when, where and how genes are expressed in order to foment both discovery and hypothesis generation. In fact, hypothesis-driven biological research solicits models to represent biological processes. Once models are created, they are tested against experimental results. The design of models and of the proper experiments allowing to effectively conclude about their validity is a crucial step in the discovery process. The availability, diversity, robustness, and correct interpretation of prior experimental results, such as those from microarray experiments, are therefore instrumental in formulating new hypotheses and models, as well as in designing the proper experiments to test them. Genevestigator-Mouse aims at providing easy-to-use but powerful tools that enable biologists to obtain context-driven information about the expression of the mouse transcriptome. The information obtained helps to validate existing hypotheses, as well as to formulate new hypotheses or to design novel experiments.
Construction and content
Data source, processing, and annotation
Data was downloaded via FTP from public repositories such as Gene Expression Omnibus , ArrayExpress , MUSC , PEPR , ChipperDB  or NIH Neuroscience Microarray Consortium . Raw data (CEL files) were normalized with the Affy package from Bioconductor  using the MAS5 algorithm. Experiment annotations were retrieved from public repositories, from original publications, and occasionally directly from the authors. Anatomy ontologies, of which 160 are currently represented in the database, are based on definitions provided by the Edinburgh Mouse Atlas Project and available at Mouse Genome Informatics . Developmental stages are partitioned into 27 pre-natal  and 5 post-natal stages. In the latter case, stages were defined based on a log(4) scale of time units (days) after birth. Genetic modifications were systematically annotated according to the underlying mutagenesis methods, e.g. targeted deletion or ENU mutagenesis and including, if possible, information about which genes were affected. As for treatments and stimuli, several major categories currently cover 80 treated samples (+) and their corresponding controls (-). Data for the mapping of probe sets to gene identifiers were obtained from the Affymetrix website . Currently, either probe set or UniGene identifiers can be used for querying the database.
A prerequisite for the type of analysis provided by Genevestigator is data comparability between experiments. Although methods how to combine data from different technological platforms and laboratories are still a matter of debate, the common analysis of data from a single organism, a single platform such as the Affymetrix system, and a single array type has so far proven to successfully reveal biological mechanisms. In fact, as can be verified from recent publications, results obtained from Genevestigator with Arabidopsis could be substantiated by other experimental techniques such as RT-PCR, reporter gene analysis, mutant phenotyping, or other microarray studies. In order to maximize comparability, quality-control (QC) measures were applied to raw and normalized data. The current QC protocol uses several Bioconductor  packages (simpleaffy, AffyQCReport, AffyPLM) and in-house R code. QC results include signal intensity box-plots and density-plots, Actin and GAPDH QC statistics, RLE, NUSE, RNA digestion plots, positive and negative border element plots, as well a correlation matrix. QC reports are publicly available. Array data containing flaws or biases, or of which RNA was strongly degraded, are flagged and excluded by default from the data analysis tools. Currently, approximately 3.4% of arrays did not fulfil the QC criteria. However, users can decide whether to use all the available data by inactivating the default "Quality Control" option.
Database and analysis tools
The database contains both 12 K and 40 K Affymetrix arrays (MG-U74Av2 and Mouse430 2.0). As of June 2006, 3110 arrays from 166 experiments were curated, quality-controlled and annotated according to controlled vocabularies. Two types of queries can be run: gene-centric (how is gene X expressed in a series of conditions?) or genome-centric (which genes are expressed in a pre-selection of conditions?). The tools described below contain either one or both types. Data can be viewed either in linear or log(2)-scale relative to the following categories:
Experiments and arrays (Digital Northern): the digital Northern visualizes signal intensity values and present/absent calls of a subset of genes across a selection of experiments.
Anatomy (Gene Atlas): this tool reveals the anatomy-specific profiles of single query genes, or reversely allows to identify genes expressed specifically in a selection of organs/tissues.
Stimulus (Response Viewer): this tool shows the responses of genes to a compendium of stimuli. Results can be sorted such as to rapidly identify those factors most affecting the expression of query genes. Reversely, it allows to find genes expressed specifically in a subset of conditions.
Development (Gene Chronologer): changes in gene expression throughout the life cycle of the mouse, which is grouped into 7 embryo and 5 post-natal stage groups, can be visualized. Reversely, one can search for genes expressed specifically at given stages of development.
Mutation (Mutant Surveyor): this tool is similar to the Response Viewer, plotting the responses of a query gene to a number of genetic backgrounds or modifications (e.g. gene knock-out or overexpression).
Multiple genes (Meta-Analyzer): expression profile summaries of a larger number of genes with respect to anatomy, development, or stimulus can be queried.
Documentation: the documentation and FAQ sections provide important technical and practical information about the tools, such as statistical procedures, probe set specificity, normalization, how to interpret data, and precautions to avoid over-interpretation.
Database: this page provides information about all experiments stored in our database, including data source repositories, authors, publications, dates, original file names, Genevestigator file names, annotations, quality-control results, and links to external information. Since Genevestigator is basically a data analysis tool and not a repository, we do not provide data bulk download options and therefore recommend users to download the data from the original repositories.
Utility and discussion
A validation and discovery study is presented to illustrate some of the querying possibilities and to assess the quality of the output by comparison to prior biological knowledge obtained from the literature. Besides confirming previously known mechanisms, we present novel findings and hypotheses about the regulation of given genes in mouse.
Our database encompasses a wide range of experimentally annotated gene expression data covering diverse categories i.e. different tissues, developmental stages, treatments and genetic modifications. Since the averages computed for each category are based on data from a variety of experimental setups that are not necessarily systematic throughout all categories, care must be taken not to over-interpret results. Our general hypothesis is that an increase in the number of experiments and replicates per defined category generally attenuates experiment specific effects in favor of global trends. To verify whether results from Genevestigator show robust and reproducible trends, we selected from the literature a number of genes with expression profiles specific to organs or to developmental stages, or genes responsive to certain stimuli or to genetic modifications and analyzed their respective expression patterns generated by Genevestigator.
Second, to verify the reliability of the Development profiles (Gene Chronologer tool), we looked for genes annotated as being developmentally regulated. Only a handful of early embryonic genes have been described to date. Among those we tested the RING finger protein gene Rnf33 [Mm.28010] and Hoxal [Mm.197] a homeobox transcription factor that regulates embryonic patterning and organogenesis. Transcription of Rnf33 has been shown to occur already in the mouse oocyte but not beyond the eight-cell stage nor in adult tissues . Hoxal expression starts at E7.5 and begins to retreat caudally by day E8.5 . Both genes were found to be expressed solely in the corresponding embryonic developmental stage groups (Fig. 1B1 and 1B2) [see Additional file 1]. In contrast, the gene encoding hemopexin (hx, [Mm.3485]), a plasma glycoprotein known to be only lowly expressed in embryos and newborn mice  showed an adult stage specific expression profile (Fig. 1B3) with the strongest signal at the latest stage, which reflects the fact that hx gene expression reaches adult level not until the first year of age .
Third, to test the Genevestigator output in terms of responses to different stimuli (Response Viewer) and to genetic modifications (Mutant Surveyor), we used Sirt1 [Mm.351459], a nuclear deacetylase that is closely associated with the longevity elicited by caloric restriction (CR, , Fig. 1C) [see Additional file 1]. In mammals a characteristic set of physiological changes takes place during CR. Among those changes is the use of dietary fat or fat mobilized from white adipose tissue for energy, and a large reduction in blood insulin levels accompanied by an increase in insulin sensitivity. It was shown that the flux of metabolites under fasting conditions leads to an increase in Sirt1 mRNA levels and that Sirt1 also controls glucose metabolism through the regulation of PPAR co-activator 1 . The Response Viewer correctly showed an up regulation of Sirt1 in fasted mice, in mice which were put on a fat diet, as well as in response to several insulin treatments. No change or even a slight down regulation in Sirt1 expression was observed in caloric restriction treatments performed with Ames mice. This is consistent with the fact that the pathways responsible for extending lifespan in the dwarfs and in CR animals are not identical . In addition, we found Sirt1 severely up regulated in response to dexamethasone, in accordance with the reported increase of Sirt1 protein levels upon dexamethasone treatment measured by western blot analysis of 3T3-L1 fibroblasts . The Mutant Surveyor analysis revealed a substantial increase of Sirt1 mRNA levels in mice with a truncated growth hormone receptor (mutant 391), which are characterized by marked male obesity associated with hyperglycemia; in BDC2.5/NOD mice, which develop a mild cellular infiltrate in the pancreatic islets of Langerhan's (insulitis); and in C57BL/6J-ob/ob mice, which exhibit a diabetes-like syndrome of hyperglycemia, glucose intolerance and elevated plasma insulin levels  (Fig. 1D). Hence, the Sirt1 expression profile in these mouse strains reflects its role in controlling glucose metabolism. Moreover, Sirt1 was found to be up regulated in PPAR over-expressing mice, which is consistent with the model that Sirt1 functions as a represser of genes that drive white adipocyte differentiation and fat storage . Remarkably, among the genetically modified strains that showed increased Sirt1 expression were also telomerase-deficient mice (Terc-/-, ), mice defective in the Ku86 DNA repair protein (Ku86-/-, ) and mice double deficient for telomerase and Ku86 (Terc-/-/Ku86-/-, ) all characterized by a premature-aging phenotype. This in turn is consistent with the observation that the yeast Sirt1 ortholog Sir2, a central player in yeast aging, represses transcription near telomeres  and that both Sir2 and Ku86 are involved in non-homologous end joining, which is used to repair breaks in DNA by ligation of the free ends .
This validation approach clearly shows that Genevestigator is capable to detect biologically relevant trends in the expression profiles of individual genes by combining numerous normalized expression data sets using the same technical platform, i.e. the Affymetrix system.
Based on our validation study, we conclude that with the current set of data, Genevestigator generates high quality results. Moreover, we expect that this quality will continue to rise as the size of the dataset increases.
Using standardized protocols and systematic annotation, biologically relevant trends in gene expression can be identified. The combined average of several measurements within an annotation category increases the confidence about signal values obtained from these categories. At the same time, the quality and power of the tools will increase as more numerous and more diverse data become available and are included. Unfortunately, several interesting published experiments could not be included owing to lacking experiment annotation, highlighting the importance of proper and sufficient annotation with controlled vocabularies and following detailed guidelines. Furthermore, the availability of raw data (CEL files) generated with the 40 K array were sparsely available in repositories. We therefore encourage authors to provide also raw data when submitting their experiments to public repositories.
Due to the compilation of data from different labs, marker identification methods may be sensitive to arrays with signals resulting from very weak hybridization or signals that are in the saturation range. For this reason, applying strict quality-control measures based on raw data analysis is imperative for robust analysis. Genevestigator can be used either as a validation tool for genes already under study, as a tool to create new hypotheses, or to identify marker genes. Our results demonstrate that Genevestigator effectively identifies biological mechanisms and meta-data related marker genes.
Availability and requirements
Project name: Genevestigator; Project homepage: https://www.genevestigator.ethz.ch; Operating system: platform independent; Programming language: PHP4/MySQL; Licence: free access to current tools for academics; Any restrictions to use by non-academics: licence needed.
We thank all scientists who provided experiment descriptions and raw data files, as well as public repositories for making this data available. Thanks also to Matthew Zapala for providing us with a large compendium of mouse Affymetrix arrays. This project was funded by ETH Zurich, Strategic Excellence Project 2-74213-02 TH-802-2, and Project 2-76663-05.
- Brazma A, Hingamp P, Quackenbush J, Sherlock G, Spellmann P, Stoeckert C, Aach J, Ansorge W, Ball C, Causton H, Gaasterland T, Glenisson P, Holstege F, Kim I, Markowitz V, Matese J, Parkinson H, Robinson A, Sarkans U, Schulze-Kremer S, Stewart J, Taylor R, Vilo J, Vingron M: Minimum information about a microarray experiment (MIAME) – toward standards for microarray data. Nat Genet 2001, 29: 365–371. 10.1038/ng1201-365View ArticlePubMedGoogle Scholar
- Zimmermann P, Schildknecht B, Craigon D, Garcia-Hernandez M, Gruissem W, May S, Mukherjee G, Parkinson H, Rhee S, Wagner U, Hennig L: MIAME/Plant – adding value to plant microarray experiments. Plant Methods 2006, 2: 1. 10.1186/1746-4811-2-1PubMed CentralView ArticlePubMedGoogle Scholar
- Bao W, Schmid J, Goetz A, Ren H, Dix D: A database for tracking toxicogenomics samples and procedures. Reprod Toxicol 2005, 19: 411–419. 10.1016/j.reprotox.2004.09.005View ArticlePubMedGoogle Scholar
- Field D, Tiwari B, Snape J: Bioinformatics and data management support for environmental genomics. PLOS Biol 2005, 3: e297. 10.1371/journal.pbio.0030297PubMed CentralView ArticlePubMedGoogle Scholar
- Psarros M, Steffen H, Sick M, Thoppae G, Harshman K, Sick B: RACE: Remote Analysis Computation for gene Expression data. Nucleic Acids Res 2005, 33: W638-W643. 10.1093/nar/gki490PubMed CentralView ArticlePubMedGoogle Scholar
- Argraves G, Jani S, Barth J, Argraves W: ArrayQuest: a web resource for the analysis of DNA microarray data. BMC Bioinformatics 2005, 6: 287. 10.1186/1471-2105-6-287PubMed CentralView ArticlePubMedGoogle Scholar
- Wille A, Zimmermann P, Vranova E, Fuerhold A, Laule O, Bleuler S, Prelic A, von Rohr P, Thiele L, Zitzler E, Gruissem W, Buehlmann P: Sparse graphical Gaussian modeling of the isoprenoid gene network in Arabidopsis thaliana. Genome Biol 2004, 5(11):R92. 10.1186/gb-2004-5-11-r92PubMed CentralView ArticlePubMedGoogle Scholar
- Zimmermann P, Hirsch-Hoffmann M, Hennig L, Gruissem W: GENEVESTIGATOR. Arabidopsis microarray database and analysis toolbox. Plant Physiol 2004, 136: 2621–2632. 10.1104/pp.104.046367PubMed CentralView ArticlePubMedGoogle Scholar
- Barrett T, Suzek T, Troup D, Wilhite S, Ngau W, Ledoux P, Rudnev D, Lash A, Fujibuchi W, Edgar R: NCBI GEO: mining millions of expression profiles – database and tools. Nucleic Acids Res 2005, 1(33):D562–566.Google Scholar
- Parkinson H, Sarkans U, Shojatalab M, Abeygunawardena N, Contrino S, Coulson R, Fame A, Lara G, Holloway E, Kapush- esky M, Lilja P, Mukherjee G, Oezcimen A, Rayner T, Rocca-Serra P, Sharma A, Sansone S, Brazma A: ArrayExpress – a public repository for microarray gene expression data at the EBI. Nucleic Acids Res 2005, 1(33):D553–555.Google Scholar
- Argraves G, Barth J, Argraves W: The MUSC DNA Microarray Database. Bioinformatics 2003, 19(18):2473–2474. 10.1093/bioinformatics/btg325View ArticlePubMedGoogle Scholar
- Chen J, Zhao P, Massaro D, Clerch L, Almon R, DuBois D, Jusko W, Hoffman E: The PEPR GeneChip data ware house, and implementation of a dynamic time series query tool (SGQT) with graphical interface. Nucleic Acid Res 2004, 32: D578–581. 10.1093/nar/gkh003PubMed CentralView ArticlePubMedGoogle Scholar
- NIH Neuroscience Microarray Consortium[http://arrayconsortium.tgen.org]
- Mouse Anatomical Dictionary[http://www.informatics.jax.org/searches/anatdict_form.shtml]
- Theiler K: The house mouse : atlas of embryonic development. New York, Springer Verlag; 1989.View ArticleGoogle Scholar
- Gentleman R, Carey V, Douglas M, Bolstad B, Dettling M, Dudoit S, Ellis B, Gautier L, Ge Y, Gentry J, Hornik K, Hothorn T, Huber W, Iacus S, Irizarry R, Leisch F, Li C, Maechler M, Rossini A, sawitzki G, Smith C, Smyth G, Tierney L, Yang Y, Zhang J: Bioconductor: open software development for computational biology and bioinformatics. Genome Biol 2004, 5: R80. 10.1186/gb-2004-5-10-r80PubMed CentralView ArticlePubMedGoogle Scholar
- Furukawa R, Morrow E, Cepko C: Crx, a novel otx-like homeobox gene, shows photoreceptor-speciflc expression and regulates photoreceptor differentiation. Cell 1997, 91: 531–541. 10.1016/S0092-8674(00)80439-0View ArticlePubMedGoogle Scholar
- Essayan D: Cyclic nucleotide phosphodiesterases. J Allergy Clin Immunol 2001, 108: 671–680. 10.1067/mai.2001.119555View ArticlePubMedGoogle Scholar
- Swaroop A, Xu J, Pawar H, Jackson A, skolnick C, Agarwal N: A conserved retina-specific gene encodes a basic motif/leucine zipper domain. Proc Natl Acad Sci USA 1992, 89(1):266–270. 10.1073/pnas.89.1.266PubMed CentralView ArticlePubMedGoogle Scholar
- Granzier H, Labeit D, Wu Y, Witt C, Watanabe K, Lahmers S, Gotthardt M, Labeit S: Adaptations in titin's spring elements in normal and scardiomyopathic hearts. Adv Exp Med Biol 2003, 538: 517–530.View ArticlePubMedGoogle Scholar
- Siebrands C, Sanger J, Sanger J: Myofibrillogenesis in skeletal muscle cells in the presence of taxol Cell Motil. Cytoskeleton 2004, 58: 39–52. 10.1002/cm.10177View ArticleGoogle Scholar
- Phan D, Rasmussen T, Nakagawa O, McAnally J, Gottlieb P, Tucker P, Richardson J, Bassel-Duby R, Olson E: BOP, a regulator of right ventricular heart development, is a direct transcriptional target of MEF2C in the developing heart. Development 2005, 132(11):2669–2678. 10.1242/dev.01849View ArticlePubMedGoogle Scholar
- Choo K, Chen H, Cheng W, Chang H, Wang M: In silico mining of EST databases for novel pre-implantation embryo-specific zinc finger protein genes. Mol Rep Dev 2001, 59: 249–255. 10.1002/mrd.1029View ArticleGoogle Scholar
- Remacle S, Abbas L, Backer O, Pacico N, Gavalas A, Gofflot F, Picard J, Rezsohazy R: Loss of function but no gain of function caused by amino acid substitutions in the hexapeptide of Hoxal in vivo. Mol Cell Biol 2004, 24(19):8567–8575. 10.1128/MCB.24.19.8567-8575.2004PubMed CentralView ArticlePubMedGoogle Scholar
- Tolosano E, Cutufia M, Hirsch E, Silengo L, Altruda F: Specific expression in brain and liver driven by the hemopexin promoter in transgenic mice. Biophys Res Commun 1996, 218(3):694–703. 10.1006/bbrc.1996.0124View ArticleGoogle Scholar
- Leibiger I, Berggren P: Sirt1: a metabolic master switch that modulates lifespan. Nat Med 2006, 12(1):34–36. 10.1038/nm0106-34View ArticlePubMedGoogle Scholar
- Rodgers J, Lerin C, Haas W, Gygi S, Spiegelman B, Puigserver P: Nutrient control of glucose homeostasis through a complex of PGC-lalpha and SIRT1. Nature 2005, 434(7029):113–118. 10.1038/nature03354View ArticlePubMedGoogle Scholar
- Bartke A, Wright C, Mattison J, Ingram D, Miller R, Roth G: Extending the lifespan of long-lived mice. Nature 2001, 414: 412. 10.1038/35106646View ArticlePubMedGoogle Scholar
- Picard F, Kurtev M, Chung N, Topark-Ngarm A, Senawong T, Machado De Oliveira R, Leid M, McBurney M, Guarente L: Sirt1 promotes fat mobilization in white adipocytes by repressing PPAR-gamma. Nature 2004, 429: 771–776. 10.1038/nature02583PubMed CentralView ArticlePubMedGoogle Scholar
- Lan H, Rabaglia M, Stoehr J, Nadler S, Schueler K, Zou F, Yandell B, Attie A: Gene expression profiles of nondiabetic and diabetic obese mice suggest a role of hepatic lipogenic capacity in diabetes susceptibility. Diabetes 2003, 52(3):688–700.View ArticlePubMedGoogle Scholar
- Blasco M, Lee H, Hande M, Samper E, Lansdorp P, DePinho R, Greider C: Telomere shortening and tumor formation by mouse cells lacking telomerase RNA. Cell 1997, 91(1):25–34. 10.1016/S0092-8674(01)80006-4View ArticlePubMedGoogle Scholar
- Vogel H, Lim D, Karsenty G, Finegold M, Hasty P: Deletion of Ku86 causes early onset of senescence in mice. Proc Natl Acad Sci USA 1999, 96(19):10770–10775. 10.1073/pnas.96.19.10770PubMed CentralView ArticlePubMedGoogle Scholar
- Espejel S, Klatt P, Menissier-de Murcia J, Martin-Caballero J, Flores J, Taccioli G, de Murcia G, Blasco M: Impact of telomerase ablation on organismal viability, aging, and tumorigenesis in mice lacking the DNA repair proteins PARP-1, Ku86, or DNA-PKcs. J Cell Biol 2004, 167: 627–638. 10.1083/jcb.200407178PubMed CentralView ArticlePubMedGoogle Scholar
- Aparicio O, Billington B, Gottschling D: Modifiers of position effect are shared between telomeric and silent mating-type loci in S. cerevisiae. Cell 1991, 66(6):1279–1287. 10.1016/0092-8674(91)90049-5View ArticlePubMedGoogle Scholar
- Kaeberlein M, McVey M, Guarente L: The SIR2/3/4 complex and SIR2 alone promote longevity in Saccha- romyces cerevisiae by two different mechanisms. Genes Dev 1999, 13(19):2570–2580. 10.1101/gad.13.19.2570PubMed CentralView ArticlePubMedGoogle Scholar
- Komuro I, Izumo S: Csx: a murine homeobox-containing gene specifically expressed in the developing heart. Proc Natl Acad Sci USA 1993, 90(17):8145–8149. 10.1073/pnas.90.17.8145PubMed CentralView ArticlePubMedGoogle Scholar
- Pan J, Hinzmann B, Yan W, Wu F, Morser J, Wu Q: Genomic structures of the human and murine corin genes and functional GATA elements in their promoters. J Biol Chem 2002, 277(41):38390–38398. 10.1074/jbc.M205686200View ArticlePubMedGoogle Scholar
- Di Lisi R, Millino C, Calabria E, Altruda F, Schiaffino S, Ausoni S: Combinatorial cis-acting elements control tissue- specific activation of the cardiac troponin I gene in vitro and in vivo. J Biol Chem 1998, 273: 25371–25380. 10.1074/jbc.273.39.25371View ArticlePubMedGoogle Scholar
- Chan K, Tsui S, Lee S, Luk S, Liew C, Fung K, Waye M, Lee C: Molecular cloning and characterization of FHL2, a novel LIM domain protein preferentially expressed in human heart. Gene 1998, 210(2):345–350. 10.1016/S0378-1119(97)00644-6View ArticlePubMedGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.