Development and validation of computational models of cellular interaction

In this paper we take the view that computational models of biological systems should satisfy two conditions – they should be able to predict function at a systems biology level, and robust techniques of validation against biological models must be available. A modelling paradigm for developing a predictive computational model of cellular interaction is described, and methods of providing robust validation against biological models are explored, followed by a consideration of software issues.     

Development and validation of computational models for mammalian circadian oscillators

Circadian rhythms are endogenous rhythms with a cycle length of approximately 24 h. Rhythmic production of specific proteins within pacemaker structures is the basis for these physiological and behavioral rhythms. Prior work on mathematical modeling of molecular circadian oscillators has focused on the fruit fly, Drosophila melanogaster. Recently, great advances have been made in our understanding of the molecular basis of circadian rhythms in mammals. Mathematical models of the mammalian circadian oscillator are needed to piece together diverse data, predict experimental results, and help us understand the clock as a whole. Our objectives are to develop mathematical models of the mammalian circadian oscillator, generate and test predictions from these models, gather information on the parameters needed for model development, integrate the molecular model with an existing model of the influence of light and rhythmicity on human performance, and make models available in BioSpice so that they can be easily used by the general community. Two new mammalian models have been developed, and experimental data are summarized. These studies have the potential to lead to new strategies for resetting the circadian clock. Manipulations of the circadian clock can be used to optimize performance by promoting alertness and physiological synchronization.     

Experimental validation of finite element predicted bone strain in the human metatarsal

Metatarsal stress fracture is a common injury observed in athletes and military personnel. Mechanical fatigue is believed to play an important role in the etiology of stress fracture, which is highly dependent on the resulting bone strain from the applied load. The purpose of this study was to validate a subject-specific finite element (FE) modeling routine for bone strain prediction in the human metatarsal. Strain gauge measurements were performed on 33 metatarsals from seven human cadaveric feet subject to cantilever bending, and subject-specific FE models were generated from computed tomography images. Material properties for the FE models were assigned using a published density-modulus relationship as well as density-modulus relationships developed from optimization techniques. The optimized relationships were developed with a ‘training set’ of metatarsals (n = 17) and cross-validated with a ‘test set’ (n = 16). The published and optimized density elasticity equations provided FE-predicted strains that were highly correlated with experimental measurements for both the training (r² ≥ 0.95) and test (r² ≥ 0.94) sets; however, the optimized equations reduced the maximum error by 10% to 20% relative to the published equation, and resulted in an X = Y type of relationship between experimental measurements and FE predictions. Using a separate optimized density-modulus equation for trabecular and cortical bone did not improve strain predictions when compared to a single equation that spanned the entire bone density range. We believe that the FE models with optimized material property assignment have a level of accuracy necessary to investigate potential interventions to minimize metatarsal strain in an effort to prevent the occurrence of stress fracture.     

Bridging experiments, models and simulations: an integrative approach to validation in computational cardiac electrophysiology

Computational models in physiology often integrate functional and structural information from a large range of spatiotemporal scales from the ionic to the whole organ level. Their sophistication raises both expectations and skepticism concerning how computational methods can improve our understanding of living organisms and also how they can reduce, replace, and refine animal experiments. A fundamental requirement to fulfill these expectations and achieve the full potential of computational physiology is a clear understanding of what models represent and how they can be validated. The present study aims at informing strategies for validation by elucidating the complex interrelations among experiments, models, and simulations in cardiac electrophysiology. We describe the processes, data, and knowledge involved in the construction of whole ventricular multiscale models of cardiac electrophysiology. Our analysis reveals that models, simulations, and experiments are intertwined, in an assemblage that is a system itself, namely the model-simulation-experiment (MSE) system. We argue that validation is part of the whole MSE system and is contingent upon 1) understanding and coping with sources of biovariability; 2) testing and developing robust techniques and tools as a prerequisite to conducting physiological investigations; 3) defining and adopting standards to facilitate the interoperability of experiments, models, and simulations; 4) and understanding physiological validation as an iterative process that contributes to defining the specific aspects of cardiac electrophysiology the MSE system targets, rather than being only an external test, and that this is driven by advances in experimental and computational methods and the combination of both.     

Identification of novel Drosophila meiotic genes recovered in a P-element screen.

The segregation of homologous chromosomes from one another is the essence of meiosis. In many organisms, accurate segregation is ensured by the formation of chiasmata resulting from crossing over. Drosophila melanogaster females use this type of recombination-based system, but they also have mechanisms for segregating achiasmate chromosomes with high fidelity. We describe a P-element mutagenesis and screen in a sensitized genetic background to detect mutations that impair meiotic chromosome pairing, recombination, or segregation. Our screen identified two new recombination-deficient mutations: mei-P22, which fully eliminates meiotic recombination, and mei-P26, which decreases meiotic exchange by 70% in a polar fashion. We also recovered an unusual allele of the ncd gene, whose wild-type product is required for proper structure and function of the meiotic spindle. However, the screen yielded primarily mutants specifically defective in the segregation of achiasmate chromosomes. Although most of these are alleles of previously undescribed genes, five were in the known genes alphaTubulin67C, CycE, push, and Trl. The five mutations in known genes produce novel phenotypes for those genes.     

Verification, validation and sensitivity studies in computational biomechanics

Computational techniques and software for the analysis of problems in mechanics have naturally moved from their origins in the traditional engineering disciplines to the study of cell, tissue and organ biomechanics. Increasingly complex models have been developed to describe and predict the mechanical behavior of such biological systems. While the availability of advanced computational tools has led to exciting research advances in the field, the utility of these models is often the subject of criticism due to inadequate model verification and validation (V&V). The objective of this review is to present the concepts of verification, validation and sensitivity studies with regard to the construction, analysis and interpretation of models in computational biomechanics. Specific examples from the field are discussed. It is hoped that this review will serve as a guide to the use of V&V principles in the field of computational biomechanics, thereby improving the peer acceptance of studies that use computational modeling techniques.     

Structure and expression of two kininogen genes in mice

Kininogens serve dual functions by forming a scaffold for the assembly of the protein complex initiating the surface-activated blood coagulation cascade and as precursors for the kinin hormones. While rats have three kininogen genes, for mice, cattle, and humans only one gene has been described. Here, we present sequence and expression data of a second mouse kininogen gene. The two genes, kininogen-I and kininogen-II, are located in close proximity on chromosome 16 in a head-to-head arrangement. In liver and kidney, both genes are expressed and for each gene three alternative splice variants are synthesized. Two of them are the expected high and low molecular weight isoforms known from all mammalian kininogens. However, for both genes also a third, hitherto unknown splice variant was detected which lacks part of the high molecular weight mRNA due to splicing from the low molecular weight donor site to alternative splice acceptor sites in exon 10. The physiological functions of the six kininogen isoforms predicted by these findings need to be determined.     

Adaptive variation in beach mice produced by two interacting pigmentation genes

Little is known about the genetic basis of ecologically important morphological variation such as the diverse color patterns of mammals. Here we identify genetic changes contributing to an adaptive difference in color pattern between two subspecies of oldfield mice (Peromyscus polionotus). One mainland subspecies has a cryptic dark brown dorsal coat, while a younger beach-dwelling subspecies has a lighter coat produced by natural selection for camouflage on pale coastal sand dunes. Using genome-wide linkage mapping, we identified three chromosomal regions (two of major and one of minor effect) associated with differences in pigmentation traits. Two candidate genes, the melanocortin-1 receptor (Mc1r) and its antagonist, the Agouti signaling protein (Agouti), map to independent regions that together are responsible for most of the difference in pigmentation between subspecies. A derived mutation in the coding region of Mc1r, rather than change in its expression level, contributes to light pigmentation. Conversely, beach mice have a derived increase in Agouti mRNA expression but no changes in protein sequence. These two genes also interact epistatically: the phenotypic effects of Mc1r are visible only in genetic backgrounds containing the derived Agouti allele. These results demonstrate that cryptic coloration can be based largely on a few interacting genes of major effect.     

Gene trap insertional mutagenesis in mice: New vectors and germ Line mutations in two novel genes

Insertional mutagenesis based on gene trap vectors that capture endogenous splice sites is a promising tool for functional genomics. Several groups have proposed large-scale gene trap screens, but questions remain as to the type of vectors and their design. We report a set of plasmid-encoded gene trap vectors and the disruption of two novel genes. Our results include a comparison of the relative gene trapping efficiencies of two different splice acceptor sequences in ES cells and an analysis of the structure of several gene trap insertions.     

Structure of two in tandem human 17 beta-hydroxysteroid dehydrogenase genes

Two human 17 beta-hydroxysteroid dehydrogenase (17 beta-HSD) genes (h17 beta-HSDI and h17 beta-HSDII) included in tandem within an approximately 13 kilobase pair fragment were isolated from a genomic lambda EMBL3 DNA library using cDNA encoding human 17 beta-HSD (hpE2DH216) as probe. We have determined the complete exon and intron sequences of the two genes as well as their 5' and 3'-flanking regions. Human 17 beta-HSDII contains six exons and five short introns for a total length of 3250 base pairs. The exon sequence of h17 beta-HSDII is identical to the previously reported hpE2DH216 cDNA while the overlapping nucleotide sequences of the corresponding exons and introns of h17 beta-HSDI and h17 beta-HSDII show 89% homology. In addition, we have used the hpE2DH216 cDNA to demonstrate the widespread expression of 17 beta-HSD mRNAs in steroidogenic and peripheral target tissues. These new findings provide the basis for a better understanding of the molecular mechanisms involved in 17 beta-HSD deficiency and peripheral sex steroid metabolism.     

Gene trapping of two novel genes, Hzf and Hhl, expressed in hematopoietic cells

Using an expression gene trapping strategy, we have identified and characterized two novel hematopoietic genes, Hzf and Hhl. Embryonic stem (ES) cells containing a gene trap vector insertion were cultured on OP9 stromal cells to induce hematopoietic differentiation and screened for lacZ reporter gene expression. Two ES clones displaying lacZ expression within hematopoietic cells in vitro were used to generate mice containing the gene trap integrations. Paralleling this in vitro expression pattern, both Hzf and Hhl were expressed in a tissue-specific manner during hematopoietic development in vivo. Hzf encodes a novel protein containing three C(2)H(2)-type zinc fingers predominantly expressed in megakaryocytes and CFU-GEMM. Hhl encodes a novel protein containing a putative phosphotyrosine binding (PTB) domain expressed in megakaryocytes, CFU-GEMM and BFU-E. These results demonstrate the utility of expression trapping to identify novel hematopoietic genes. Future studies of Hzf and Hhl should provide valuable information on the role these genes play during megakaryocytopoiesis.     

Experimental validation of in silico model‐predicted isocitrate dehydrogenase and phosphomannose isomerase from Dehalococcoides mccartyi

Gene sequences annotated as proteins of unknown or non‐specific function and hypothetical proteins account for a large fraction of most genomes. In the strictly anaerobic and organohalide respiring Dehalococcoides mccartyi, this lack of annotation plagues almost half the genome. Using a combination of bioinformatics analyses and genome‐wide metabolic modelling, new or more specific annotations were proposed for about 80 of these poorly annotated genes in previous investigations of D. mccartyi metabolism. Herein, we report the experimental validation of the proposed reannotations for two such genes (KB1_0495 and KB1_0553) from D. mccartyi strains in the KB‐1 community. KB1_0495 or DmIDH was originally annotated as an NAD⁺‐dependent isocitrate dehydrogenase, but biochemical assays revealed its activity primarily with NADP⁺ as a cofactor. KB1_0553, also denoted as DmPMI, was originally annotated as a hypothetical protein/sugar isomerase domain protein. We previously proposed that it was a bifunctional phosphoglucose isomerase/phosphomannose isomerase, but only phosphomannose isomerase activity was identified and confirmed experimentally. Further bioinformatics analyses of these two protein sequences suggest their affiliation to potentially novel enzyme families within their respective larger enzyme super families.     

Experimental validation of intact and implanted distal femur finite element models

Four finite element (FE) models of intact and distal femur of knee replacements were validated relative to measured bone strains. FE models of linear tetrahedrons were used. Femoral replacements with cemented stemless, cemented and noncemented femoral stems of the PFC Sigma Modular Knee System were analyzed. Bone strains were recorded at ten locations on the cortex. The magnitude of the FE bone strains corresponded to the mean measured strains, with an overall agreement of 10%. Linear regression between the FE and mean experimental strains produced slopes between 0.94 and 1.06 and R(2) values between 0.92 and 0.99. RSME values were less than 12%. The FE models were able to adequately replicate the mechanical behavior of distal femur reconstructions.