A comparative analysis of protein targets of withdrawn cardiovascular drugs in human and mouse
© Zhao et al.; licensee BioMed Central Ltd. 2012
Received: 23 December 2011
Accepted: 1 May 2012
Published: 1 May 2012
Mouse is widely used in animal testing of cardiovascular disease. However, a large number of cardiovascular drugs that have been experimentally proved to work well on mouse were withdrawn because they caused adverse side effects in human.
In this study, we investigate whether binding patterns of withdrawn cardiovascular drugs are conserved between mouse and human through computational dockings and molecular dynamic simulations. In addition, we also measured the level of conservation of gene expression patterns of the drug targets and their interacting partners by analyzing the microarray data.
The results show that target proteins of withdrawn cardiovascular drugs are functionally conserved between human and mouse. However, all the binding patterns of withdrawn drugs we retrieved show striking difference due to sequence divergence in drug-binding pocket, mainly through loss or gain of hydrogen bond donors and distinct drug-binding pockets. The binding affinities of withdrawn drugs to their receptors tend to be reduced from mouse to human. In contrast, the FDA-approved and best-selling drugs are little affected.
Our analysis suggests that sequence divergence in drug-binding pocket may be a reasonable explanation for the discrepancy of drug effects between animal models and human.
KeywordsWithdrawn cardiovascular drugs Animal modeling Sequence divergence Side effects Drug-binding pocket
Mouse is the most commonly used vertebrate species in animal testing which is a vital part of drug development. Currently, all new pharmaceuticals undergo rigorous animal testing before being licensed for human use. It is commonly accepted that animal testing on mouse can provide us critical information for assessing hazard and risk potential [1, 2] . However, many withdrawn pharmaceuticals worked very well in animal models while led to adverse side effects or failed to reach the expected effects in human. Over sixty drugs ever approved by the FDA were withdrawn in the past twenty years . A recent and well-known example is the failure of an Alzheimer’s drug Dimebon (latrepirdine). The drug improved the performance of memory-impaired mice and rats [4, 5], but failed to show a significant effect in the phase III clinical trial. Inevitably, a question arises as to why the discrepancy of drug effects between animal models and human. Answering this question will help us avoid drug withdrawals as much as possible and ‘rescue’ some of the withdrawn drugs by promoting genotype-based prescribing.
We presumed that two aspects would be the prime determinants of the discrepancy of drug response between animal models and humans. First of all, target proteins of withdrawn drugs may be functionally divergent between human and mouse. Phenotypic differences between species might be caused by changes in species-specific interactions or gene expression [6–8], which also lead to the differential responses to the exogenous drugs. Secondly, three-dimensional structures of drug targets may have changed due to sequence divergence in the drug-binding pockets. As a result, although the drug targets are functionally conserved, the binding patterns of drugs with these targets may change from animal models to human.
In this study, both genetic aspects were examined to elucidate the mechanism of discrepancy of withdrawn drugs between human and mouse. We focused on withdrawn cardiovascular drugs, not only because cardiovascular drugs are in a position of importance in all pharmaceuticals, but also because the relevant information is comprehensive, including target, pharmacodynamics and so on. The results show that the drug targets are not involved in species-specific interactions. The binding patterns of withdrawn cardiovascular drugs are indeed affected by sequence divergence in the drug-binding pockets. The trend of binding affinity of these withdrawn drugs is to be reduced from mouse to human. Our study gleans valuable insights into the development of cardiovascular drugs and may reduce the amount of laboratory effort required in this process.
Target proteins of withdrawn cardiovascular drugs are functionally conserved between human and mouse
In order to compare the target protein of withdrawn cardiovascular drugs between human and mouse, we first examined the orthologous relationships of these proteins through Ensembl gene orthology prediction pipeline (http://www.ensembl.org/). It shows that all of the ortholog relationships (25/25) are one to one (Additional file 1: Table S1), which are widely assumed to have similar functions and cause similar phenotypes .
In addition, we reasoned that changes in gene expression underlie many of the phenotypic differences between species and lead to differential responses to withdrawn drugs. As a result, we also measured the level of conservation of gene expression patterns of the drug targets and their interacting partners (Materials and methods). For the expression profiles of these orthologous genes between mouse and human, are the average Pearson’s correlation coefficient r is 0.27 and is significantly higher than that for the random gene pairs (Student’s two sample t-test; p <10−8). It shows that the gene expression evolution of these drug targets between human and mouse was strongly shaped by purifying selection, suggesting that gene expression variation may not have an influence on the drug effect.
Sequence divergence near the functional sites
We reasoned that non-conserved substitutions of critical residues in drug targets could lead to different response to drugs. As a result, whether these residues had changed from mouse to human was analyzed. The results shows that the critical sites of drug targets, such as active sites and phosphorylated residues, are conserved while residues nearby have changed a lot from mouse to human, including losses of hydrogen bond donors and substitutions of amino acid residues of opposite charge (Additional file 1: SI Table S1). For example, human plasminogen activator inhibitor 1, the target of troglitazone (Drug Bank ID: DB00197), shares only 78% sequence identity with its mouse ortholog. The critical sites are Arg369 and Met370 , which are conserved, while the residues within 10 Å distance in three-dimensional structure have changed a lot, such as Lys311Gly, His339Ser, and Glu373Thr. Previous work revealed that drug-binding sites on proteins usually exist out of functional necessity , so we can infer that the binding pockets of withdrawn drugs could be affected by sequence divergence.
Structural modeling of mouse drug targets and refinement by molecular dynamics simulation
The root mean-square fluctuations (RMSF) of the mutated residues in ten drug targets were also computed to examine the impact of mutating these residues on conformational changes. The average RMSF of the mutated residues is significant higher than that of the conserved residues in 2 ns (Student’s two-sample t-test, p = 3.52 × 10−7), indicating that mutated residues have a great influence on conformational changes of drug targets.
Differences in the binding patterns of withdrawn cardiovascular drugs to their targets between human and mouse
We next analyzed whether the drugs might bind human and mouse drug targets with different patterns and affinities. These drugs belong to different categories (Additional file 1: Table S1) and so are demonstrated separately.
Encainide can mediate the voltage-dependent sodium ion permeability of excitable membrane by blocking sodium channel protein type 5 subunit alpha (HH1) . Two residue substitutions (Leu1812Ser and Ala1813Val) occurred in the Encainide binding pocket (Additional file 1: Figure S2, A-B). The amino acid residue Ser 1812 acts as a hydrogen donor according to the identification result of drug-binding sites. However, the mouse and human HH1 show similar encainide binding affinity. (Student’s two-sample t-test, p = 0.64).
The binding pattern and binding affinity is similar for FDA-approved drugs
We next inspected that whether sequence divergence affects the binding pattern and binding affinity of FDA-approved drugs. In comparison, the FDA-approved and best-selling drugs, atorvastatin and clopidogrel (brand names are lipitor and plavix separately) were estimated. The results show that the binding patterns and binding affinities of both drugs are little affected by sequence divergence between human and mouse (SI Text Results).
In this study, we demonstrate a reasonable possibility of the discrepancy of withdrawn cardiovascular drugs between mouse model and human. The drug targets are proved to be not involved in species-specific interactions. Then, reasonable models for mouse protein structures are generated through molecular dynamics simulations. After that, we determine the differences of binding patterns and affinities of the withdrawn cardiovascular drugs between human and mouse and two FDA-approved and best-selling drugs are selected as controls. Our results show that the binding patterns of withdrawn cardiovascular drugs in our study are indeed affected by sequence divergence, especially the non-conserved residue substitutions, such as Lys into Glu, Ser into Ala and so on. The trend of binding affinity of these withdrawn drugs is to be reduced from mouse to human, which may explain the low specificity of human targets. Finally, we explore whether off-target effects could be caused. It shows that more off-target effects can occur in human than in mouse due to the local structural similarities of drug-binding sites.
A key issue in translational medicine is the need to devote more attention to the characteristics of the drug–target interaction , yet most of target proteins in animal models have not or fragmentally solved. Besides, according to previous works, one missense mutation is sufficient to lead to altered binding pattern of drug targets [24, 25]. Studies on structural divergence of drug targets between disease models and human provide valuable information for the development of human disease models. In this study, we employed MD simulations to construct models of good quality as target structures and then applied them to comparison of drug-target interactions. It will be very useful to identify a suitable model for animal testing, in case of false positive results supported by inappropriate experiments.
Although we highlight the contributions of sequence divergence, we also illustrate specific interactions that were experimentally confirmed in just one species. Recent studies have shown that species-specific interactions and regulations exist in human and mouse [26, 27]. These species-specific properties may lead to different responses to the same pharmaceutical. Besides, false positive results in mouse may also lead to the success of animal testing while failure of clinical treatments. For example, prenylamine is a calcium channel blocker of the amphetamine chemical class which is used as a vasodilator in the treatment of angina pectoris. It has been shown to partially metabolize to amphetamine and can cause false positives for it in drug tests [28–30].
In summary, our results provide a reasonable explanation for the discrepancy of cardiovascular drugs. We will further testify for other pharmaceuticals.
Materials and methods
Collection of withdrawn and FDA-approved cardiovascular drugs
The information of withdrawn drug for CADs were retrieved from DrugBank database  and previous works . Seven withdrawn cardiovascular drugs (Drug Bank ID: DB00197, DB00439, DB01228, DB01388, DB04825, DB04831 and DB04898) were included. In this dataset, prenylamine (Drug Bank ID: DB04825) was withdrawn for it partially metabolized to amphetamine and could cause false positives for it in drug tests . It was just a prodrug, so it was excluded out of our study. Ticrynafen (Drug Bank ID: DB04831) was also excluded because it had no target information. We also chose two FDA-approved and the most sales-potential drugs, Lipitor (Drug Bank ID: DB01076) and Plavix (Drug Bank ID: DB00758) as positive control because the higher sales-potential might indicate the more satisfactory efficacy and less side-effect.
Reconstruction of interaction networks of withdrawn drug targets and literature search
The interaction data was retrieved from several common databases, including DIP , BIND , HPRD , BioGRID , MINT  and IntAct . A Cytoscape  plugin, BisoGenet, was applied to reconstruct interaction networks of human and mouse drug targets separately. Only the interactions that were verified by experiments were displayed. Another cytoscape plugin, Agilent Literature Search 2.71, was used to do literature search. Gene aliases and context were set to restrain the search results.
Analysis of gene expression data
The microarray datasets (GSE8000 for mouse  and GSE11560 for human ) were downloaded from the Gene Expression Omnibus (GEO) (http://www.ncbi.nlm.nih.gov/geo). The expression profile of the drug targets were analyzed according to our previous work .
Preparation of structures of drug targets
All the protein structure files of withdrawn drug targets were downloaded from PDB database . Structures, containing the main functional domains and of which sequence coverage is above 80%, were adopted to do the subsequent analysis. If the sequence coverage was below the cutoff value and yet the drug-binding regions were intact and clearly reported in a structure, then the structure was also included in the dataset (Additional file 1: Table S1).
Generation of mutant models for mouse targets and refinement
Mutant models of human protein structures were made as mouse protein structures using MODELER program.  High levels of molecular dynamics with simulated annealing were performed to optimize the local structures of mutant models. After that, molecular dynamics simulations were performed using NAMD and CHARMM31 force filed with CMAP correction [44, 45]. The protein models were solvated in a TIP3P water box  and ionized by NaCl (0.152 M) to mimic physiological conditions. The ionized systems were minimized for 50,000 integration steps and equilibrated for 2 ns with 2 fs time stepping. Following this, a 2 ns unconstrained equilibration was performed for subsequent trajectory analysis, with frames stored each picosecond. Constant temperature (T = 310 K) was enforced using Langevin dynamics with a damping coefficient of 5 ps. Constant pressure (p = 1 atm) was enforced through the Nosé-Hoover Langevin piston method with a decay period of 100 fs and a damping time constant of 5 per picosecond. Van der Waals interaction cutoff distances were set at 12 Å, (smooth switching function beginning at 10 Å) and long-range electrostatic forces were computed using the particle-mesh Ewald (PME) with a grid size of less than 1.0 Å.
Comparison of binding patterns of cardiovascular drugs
We compared modes of actions of withdrawn cardiovascular drugs following the steps below. Step one; we detect the drug binding pockets through a CHARMm-based molecular dynamics scheme . We performed simulated annealing refinement and saved the top ten poses to prepare for the calculation of binding energy. Step two: a flexible docking method was applied to find the optimal binding sites by simulating protein flexibility and docking drugs with an induced fit receptor optimization . Step three; we determined the critical drug-binding sites by using Receptor-Ligand Interactions tool. We set 2.3 Å as cutoff value to exclude the minor important bonds. All simulations reported above were carried out using Discovery Studio© v126.96.36.19964, built on the SciTegic Enterprise Server platform (Accelrys Inc, 2009).
Detection of local structural similarities of drug-binding sites in structure databases
An accurate algorithm for detecting local protein structural similarity, CMASA  was applied to detect local structural similarities of drug-binding sites in non-redundant structure database from SCOP database . The binding sites detected above were used as queries. Substitution files were applied in the process and the cutoff was set to the default value.
Binding free energy calculations and statistical analysis
We used the top ten poses mentioned above to calculate binding energies. In situ ligand minimization and ligand conformational entropy were set to default value. The distance cutoff value of ligand conformational entropy was set to 14.0 Å. If the ten calculated binding energies (Additional file 1: Table S3) followed a normal distribution, Student’s two-sample t test was applied to compare the binding affinity of drugs to their targets between human and mouse.
Yuqi Zhao conceived, designed and performed the experiments. Jingwen Wang and Yanjie Wang analyzed the results and revised the manuscript. Jingfei Huang supervised the work and wrote the manuscript with support from all authors. All authors read and approved the final manuscript.
This work was supported by the National Natural Science Foundation of China (Grant No. 31123005), Chinese Academy of Sciences (Grant No. Y002731071) and the National Basic Research Program of China (Grant Nos. 2009CB941300).
- Baker M: Animal models: inside the minds of mice and men. Nature. 2011, 475: 123-128. 10.1038/475123a.View ArticlePubMedGoogle Scholar
- Dolgin E: Animal testing alternatives come alive in US. Nat Med. 2010, 16: 1348-1348. 10.1038/nm1210-1348.View ArticlePubMedGoogle Scholar
- Shah RR: Can pharmacogenetics help rescue drugs withdrawn from the market?. Pharmacogenomics. 2006, 7: 889-908. 10.2217/146224188.8.131.529.View ArticlePubMedGoogle Scholar
- Miller G: PHARMACOLOGY The Puzzling Rise and Fall of a Dark-Horse Alzheimer’s Drug. Science. 2010, 327: 1309-1309. 10.1126/science.327.5971.1309.View ArticlePubMedGoogle Scholar
- Wang J, Ferruzzi MG, Varghese M, Qian XJ, Cheng A, Xie M, Zhao W, Ho L, Pasinetti GM: Preclinical study of dimebon on beta-amyloid-mediated neuropathology in Alzheimer’s disease. Molecular Neurodegeneration. 2011, 6: 7-10.1186/1750-1326-6-7.PubMed CentralView ArticlePubMedGoogle Scholar
- Benfey PN, Mitchell-Olds T: Perspective - From genotype to phenotype: Systems biology meets natural variation. Science. 2008, 320: 495-497. 10.1126/science.1153716.PubMed CentralView ArticlePubMedGoogle Scholar
- Miller JA, Horvath S, Geschwind DH: Divergence of human and mouse brain transcriptome highlights Alzheimer disease pathways. Proceedings of the National Academy of Sciences of the United States of America. 2010, 107: 12698-12703. 10.1073/pnas.0914257107.PubMed CentralView ArticlePubMedGoogle Scholar
- Brawand D, Soumillon M, Necsulea A, Julien P, Csardi G, Harrigan P, Weier M, Liechti A, Aximu-Petri A, Kircher M: The evolution of gene expression levels in mammalian organs. Nature. 2011, 478: 343-+-View ArticlePubMedGoogle Scholar
- Liao BY, Zhang JZ: Null mutations in human and mouse orthologs frequently result in different phenotypes. Proceedings of the National Academy of Sciences of the United States of America. 2008, 105: 6987-6992. 10.1073/pnas.0800387105.PubMed CentralView ArticlePubMedGoogle Scholar
- Shoemaker BA, Panchenko AR: Deciphering protein-protein interactions. Part II. Computational methods to predict protein and domain interaction partners. Plos Comput Biol. 2007, 3: 595-601.Google Scholar
- Andreasen PA, Riccio A, Welinder KG, Douglas R, Sartorio R, Nielsen LS, Oppenheimer C, Blasi F, Dano K: Plasminogen-Activator Inhibitor Type-1 - Reactive Center and Amino-Terminal Heterogeneity Determined by Protein and Cdna Sequencing. Febs Lett. 1986, 209: 213-218. 10.1016/0014-5793(86)81113-9.View ArticlePubMedGoogle Scholar
- Hopkins AL, Groom CR: The druggable genome. Nat Rev Drug Discov. 2002, 1: 727-730. 10.1038/nrd892.View ArticlePubMedGoogle Scholar
- Fan H, Mark AE: Refinement of homology-based protein structures by molecular dynamics simulation techniques. Protein Sci. 2004, 13: 211-220. 10.1110/ps.03381404.PubMed CentralView ArticlePubMedGoogle Scholar
- Aljada A, Garg R, Ghanim H, Mohanty P, Hamouda W, Assian E, Dandona P: Nuclear factor-kappa B suppressive and inhibitor-kappa B stimulatory effects of troglitazone in obese patients with type 2 diabetes: Evidence of an antiinflammatory action?. J Clin Endocr Metab. 2001, 86: 3250-3256. 10.1210/jc.86.7.3250.PubMedGoogle Scholar
- Wang Y, Fang F, Wong CW: Troglitazone is an estrogen-related receptor alpha and gamma inverse agonist. Biochem Pharmacol. 2010, 80: 80-85. 10.1016/j.bcp.2010.03.013.View ArticlePubMedGoogle Scholar
- Ganne F, Vasse M, Beaudeux JL, Peynet J, Francois A, Mishal Z, Chartier A, Tobelem G, Vannier JP, Soria J, Soria C: Cerivastatin, an inhibitor of HMG-CoA reductase, inhibits urokinase/urokinase-receptor expression and MMP-9 secretion by peripheral blood monocytes - A possible protective mechanism against atherothrombosis. Thromb Haemostasis. 2000, 84: 680-688.Google Scholar
- Istvan ES, Palnitkar M, Buchanan SK, Deisenhofer J: Crystal structure of the catalytic portion of human HMG-CoA reductase: insights into regulation of activity and catalysis. Embo Journal. 2000, 19: 819-830. 10.1093/emboj/19.5.819.PubMed CentralView ArticlePubMedGoogle Scholar
- Guo J, Zhan S, Somers J, Westenbroek RE, Catterall WA, Roach DE, Sheldon RS, Lees-Miller JP, Li P, Shimoni Y, Duff HJ: Decrease in density of I-Na is in the common final pathway to heart block in murine hearts overexpressing calcineurin. Am J Physiol-Heart C. 2006, 291: H2669-H2679. 10.1152/ajpheart.01247.2005.View ArticleGoogle Scholar
- Mocanu MM, Gadgil S, Yellon DM, Baxter GF: Mibefradil, a T-type and L-type calcium channel blocker, limits infarct size through a glibenclamide-sensitive mechanism. Cardiovasc Drug Ther. 1999, 13: 115-122. 10.1023/A:1007732025184.View ArticleGoogle Scholar
- Monnier N, Procaccio V, Stieglitz P, Lunardi J: Malignant-hyperthermia susceptibility is associated with a mutation of the alpha(1)-subunit of the human dihydropyridine-sensitive L-type voltage-dependent calcium-channel receptor in skeletal muscle. Am J Hum Genet. 1997, 60: 1316-1325. 10.1086/515454.PubMed CentralView ArticlePubMedGoogle Scholar
- Testa L, Bhindi R, Agostoni P, Abbate A, Zoccai GGLB, van Gaal WJ: The direct thrombin inhibitor ximelagatran/melagatran: a systematic review on clinical applications and an evidence based assessment of risk benefit profile. Expert Opin Drug Saf. 2007, 6: 397-406. 10.1517/147403184.108.40.2067.View ArticlePubMedGoogle Scholar
- Iyengar R, Boran ADW: Systems approaches to polypharmacology and drug discovery. Curr Opin Drug Disc. 2010, 13: 297-309.Google Scholar
- Feuerstein GZ, Zaleska MM, Krams M, Wang X, Day M, Rutkowski JL, Finklestein SP, Pangalos MN, Poole M, Stiles GL, et al: Missing steps in the STAIR case: a Translational Medicine perspective on the development of NXY-059 for treatment of acute ischemic stroke. J Cerebr Blood F Met. 2008, 28: 217-219. 10.1038/sj.jcbfm.9600516.View ArticleGoogle Scholar
- Le L, Lee EH, Hardy DJ, Truong TN, Schulten K: Molecular Dynamics Simulations Suggest that Electrostatic Funnel Directs Binding of Tamiflu to Influenza N1 Neuraminidases. Plos Comput Biol. 2010, 6 (9): pii:e1000939-View ArticleGoogle Scholar
- Sun H, Jiang YJ, Yu QS, Luo CC, Zou JW: Effect of mutation K85R on GSK-3 beta: Molecular dynamics simulation. Biochem Bioph Res Co. 2008, 377: 962-965. 10.1016/j.bbrc.2008.10.096.View ArticleGoogle Scholar
- Heinz S, Haehnel V, Karaghiosoff M, Schwarzfischer L, Muller M, Krause SW, Rehli M: Species-specific regulation of Toll-like receptor 3 genes in men and mice. J Biol Chem. 2003, 278: 21502-21509. 10.1074/jbc.M301476200.View ArticlePubMedGoogle Scholar
- Fraenkel E, Odom DT, Dowell RD, Jacobsen ES, Gordon W, Danford TW, MacIsaac KD, Rolfe PA, Conboy CM, Gifford DK: Tissue-specific transcriptional regulation has diverged significantly between human and mouse. Nat Genet. 2007, 39: 730-732. 10.1038/ng2047.PubMed CentralView ArticlePubMedGoogle Scholar
- Musshoff F: Illegal or legitimate use? Precursor compounds to amphetamine and methamphetamine. Drug Metab Rev. 2000, 32: 15-44. 10.1081/DMR-100100562.View ArticlePubMedGoogle Scholar
- Cody JT: Precursor medications as a source of methamphetamine and/or amphetamine positive drug testing results. J Occup Environ Med. 2002, 44: 435-450. 10.1097/00043764-200205000-00012.View ArticlePubMedGoogle Scholar
- Kraemer T, Roditis SK, Peters FT, Maurer HH: Amphetamine concentrations in human urine following single-dose administration of the calcium antagonist prenylamine - Studies using fluorescence polarization immunoassay (FPIA) and GC-MS. J Anal Toxicol. 2003, 27: 68-73.View ArticlePubMedGoogle Scholar
- Wishart DS, Knox C, Guo AC, Cheng D, Shrivastava S, Tzur D, Gautam B, Hassanali M: DrugBank: a knowledgebase for drugs, drug actions and drug targets. Nucleic Acids Res. 2008, 36: D901-D906.PubMed CentralView ArticlePubMedGoogle Scholar
- Salwinski L, Miller CS, Smith AJ, Pettit FK, Bowie JU, Eisenberg D: The Database of Interacting Proteins: 2004 update. Nucleic Acids Res. 2004, 32: D449-D451. 10.1093/nar/gkh086.PubMed CentralView ArticlePubMedGoogle Scholar
- Alfarano C, Andrade CE, Anthony K, Bahroos N, Bajec M, Bantoft K, Betel D, Bobechko B, Boutilier K, Burgess E, et al: The Biomolecular Interaction Network Database and related tools 2005 update. Nucleic Acids Res. 2005, 33: D418-D424.PubMed CentralView ArticlePubMedGoogle Scholar
- Pandey A, Mishra GR, Suresh M, Kumaran K, Kannabiran N, Suresh S, Bala P, Shivakumar K, Anuradha N, Reddy R, et al: Human protein reference database - 2006 update. Nucleic Acids Res. 2006, 34: D411-D414. 10.1093/nar/gkj141.PubMed CentralView ArticlePubMedGoogle Scholar
- Tyers M, Stark C, Breitkreutz BJ, Reguly T, Boucher L, Breitkreutz A: BioGRID: a general repository for interaction datasets. Nucleic Acids Res. 2006, 34: D535-D539. 10.1093/nar/gkj109.PubMed CentralView ArticlePubMedGoogle Scholar
- Zanzoni A, Montecchi-Palazzi L, Quondam M, Ausiello G, Helmer-Citterich M, Cesareni G: MINT: a Molecular INTeraction database. Febs Lett. 2002, 513: 135-140. 10.1016/S0014-5793(01)03293-8.View ArticlePubMedGoogle Scholar
- Kerrien S, Alam-Faruque Y, Aranda B, Bancarz I, Bridge A, Derow C, Dimmer E, Feuermann M, Friedrichsen A, Huntley R, et al: IntAct - open source resource for molecular interaction data. Nucleic Acids Res. 2007, 35: D561-D565. 10.1093/nar/gkl958.PubMed CentralView ArticlePubMedGoogle Scholar
- Shannon P, Markiel A, Ozier O, Baliga NS, Wang JT, Ramage D, Amin N, Schwikowski B, Ideker T: Cytoscape: A software environment for integrated models of biomolecular interaction networks. Genome Res. 2003, 13: 2498-2504. 10.1101/gr.1239303.PubMed CentralView ArticlePubMedGoogle Scholar
- Muchir A, Pavlidis P, Decostre V, Herron AJ, Arimura T, Bonne G, Worman HJ: Activation of MAPK pathways links LMNA mutations to cardiomyopathy in Emery-Dreifuss muscular dystrophy. Journal of Clinical Investigation. 2007, 117: 1282-1293. 10.1172/JCI29042.PubMed CentralView ArticlePubMedGoogle Scholar
- Blekhman R, Oshlack A, Chabot AE, Smyth GK,, Gilad Y: Gene Regulation in Primates Evolves under Tissue-Specific Selection Pressures. Plos Genet. 2008, 4 (11): e1000271-10.1371/journal.pgen.1000271.PubMed CentralView ArticlePubMedGoogle Scholar
- Zhao Y, Sheng Z, Huang J: A systematic analysis of heart transcriptome highlights divergent cardiovascular disease pathways between animal models and humans. Mol Biosyst. 2012, 8: 504-510. 10.1039/c1mb05415e.View ArticlePubMedGoogle Scholar
- Berman HM, Westbrook J, Feng Z, Gilliland G, Bhat TN, Weissig H, Shindyalov IN, Bourne PE: The Protein Data Bank. Nucleic Acids Res. 2000, 28: 235-242. 10.1093/nar/28.1.235.PubMed CentralView ArticlePubMedGoogle Scholar
- Sali A, Blundell TL: Comparative Protein Modeling by Satisfaction of Spatial Restraints. J Mol Biol. 1993, 234: 779-815. 10.1006/jmbi.1993.1626.View ArticlePubMedGoogle Scholar
- Mackerell AD, Feig M, Brooks CL: Extending the treatment of backbone energetics in protein force fields: Limitations of gas-phase quantum mechanics in reproducing protein conformational distributions in molecular dynamics simulations. J Comput Chem. 2004, 25: 1400-1415. 10.1002/jcc.20065.View ArticlePubMedGoogle Scholar
- Buck M, Bouguet-Bonnet S, Pastor RW, MacKerell AD: Importance of the CMAP correction to the CHARMM22 protein force field: Dynamics of hen lysozyme. Biophys J. 2006, 90: L36-L38. 10.1529/biophysj.105.078154.PubMed CentralView ArticlePubMedGoogle Scholar
- Jorgensen WL, Chandrasekhar J, Madura JD, Impey RW, Klein ML: Comparison of Simple Potential Functions for Simulating Liquid Water. J Chem Phys. 1983, 79: 926-935. 10.1063/1.445869.View ArticleGoogle Scholar
- Wu GS, Robertson DH, Brooks CL, Vieth M: Detailed analysis of grid-based molecular docking: A case study of CDOCKER - A CHARMm-based MD docking algorithm. J Comput Chem. 2003, 24: 1549-1562. 10.1002/jcc.10306.View ArticlePubMedGoogle Scholar
- Koska J, Spassov VZ, Maynard AJ, Yan L, Austin N, Flook PK, Venkatachalam CM: Fully Automated Molecular Mechanics Based Induced Fit Protein-Ligand Docking Method. J Chem Inf Model. 2008, 48: 1965-1973. 10.1021/ci800081s.View ArticlePubMedGoogle Scholar
- Huang JF, Li GH: CMASA: an accurate algorithm for detecting local protein structural similarity and its application to enzyme catalytic site annotation. Bmc Bioinformatics. 2010, 11: 439.Google Scholar
- Andreeva A, Howorth D, Brenner SE, Hubbard TJP, Chothia C, Murzin AG: SCOP database in 2004: refinements integrate structure and sequence family data. Nucleic Acids Res. 2004, 32: D226-D229. 10.1093/nar/gkh039.PubMed CentralView ArticlePubMedGoogle Scholar
- Jorgensen WL, Tirado-Rives J: Contribution of conformer focusing to the uncertainty in predicting free energies for protein-ligand binding. J Med Chem. 2006, 49: 5880-5884. 10.1021/jm060763i.View 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.