Project 1: Disease severity during ciral co-infection
Project Team: Tanya Miura (Project Director), Craig Miller (Collaborator), Onesmo Balemba (Collaborator), Jake Ferguson (Post Doc), Jagdish Patel (Post Doc), JT Van Leuven (Post Doc ), Andres (Andy) Gonzalez (Graduate Student)
Viral infections in the lower respiratory tract cause severe disease and are responsible for a majority of pediatric hospitalizations, approximately 20% of which are infected by more than one viral pathogen. Clinical data indicate that disease severity can be enhanced, reduced or be unaffected by viral co-infection. However, it is not clear how unrelated viruses interact within the context of a complex host to determine disease severity. The long-term goal of this research is to uncover the causal relationships between co-infection and the resulting respiratory disease severity. Variables that will potentially predict disease severity include viral strains, doses, timing, viral competition, genetic variation in the host, and the immune response. The proposed research will develop a murine model with cellular and organismal components and a human in vitro model to test the central hypothesis that respiratory viral co-infections change disease severity both by direct viral interactions and by modulating host responses. Statistical and stochastic modeling will reveal the complex interactions between heterologous viruses within their shared target cells and host organisms. Research reported in this publication was supported by the National Institute of General Medical Sciences of the National Institutes of Health under Award Number P20GM104420. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
Project 2: Multi-level dynamics of viral co-infection
Project Team: Christine Parent (Project Director), Tanya Miura (Collaborator), Jake Ferguson (Post Doc), Andrea Gonzalez-Gonzalez (Post Doc), Jagdish Patel (Post Doc), JT Van Leuven (Post Doc), Ashley DeAguero (Technician)
With the increasing global mobility of human populations, individuals are being exposed to an increasing diversity of viruses. Many approaches are used to study viral coinfections at different organizational levels, ranging from very detailed molecular studies of specific viruses to epidemiological studies at the population level, but very few systems offer the possibility to study the multi-level dynamics of viral coinfection, from molecules to communities.
Our research is building on the host-virus system of Drosophila and associated viruses to leverage the advantages of studying large host and viral populations, powerful genetic tools, and ready access to sequencing technology. In collaboration with the CMCI, we will focus on questions typically not easily addressed in other experimental systems. Specifically, we will establish a tractable invertebrate model of viral infection and co-infection, and develop mathematical models to understand how viruses interact with each other and their host to ultimately affect the host pathology and population dynamics. Research reported in this publication was supported by the National Institute of General Medical Sciences of the National Institutes of Health under Award Number P20GM104420. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
Project 3: Agent-based modeling of co-infection
Project Team: Bert Baumgaertner (Project Director), Joseph DeAguero (Graduate Student)
How pathogens spread through a population can be complicated by a number of factors. One of them is pathogen interaction during co-infection. Here infection by one pathogen can change host susceptibility to a second, or being co-infected can change a host’s infectivity compared to a singly infected host. A second factor is that infection can alter behavior both for biological reasons—for example, when sickness makes a host-less active—and, in humans especially, for social reasons—when, for example, sick people self-isolate. These behavioral responses, in turn, change the patterns of interactions that drive transmission dynamics. A third closely related factor is that patterns of spatial aggregation around environmental features like food and water, or for humans, institutions like schools and home, can create an intricate network of interactions that strongly affect how infections spread. This project will focus on how the transmission dynamics of the population are affected by the interactions of co-infecting pathogens, the environment, and social factors that influence behavior. The main approach to this research makes use of agent-based modeling, a computational framework comprised of individuals, an environment, and rules for how individuals interact with the environment and each other. Research reported in this publication was supported by the National Institute of General Medical Sciences of the National Institutes of Health under Award Number P20GM104420. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
Project Team: Marty Ytreberg (Project Director), Jagdish Patel (Post Doc), Chris Mirabzadeh (Graduate Student)
A key to understanding protein evolution is the ability to predict how amino acid changes modify the binding strength (called affinity) between two proteins. There is a trade-off between accuracy and speed when using molecular modeling to calculate binding affinities. Our interest is to develop and test methods that have reasonable accuracy and yet are fast enough to calculate multiple affinity values per day on a single CPU.
Modeling Access Grants
Modeling Access Grants are supported by the National Institute of General Medical Sciences of the National Institutes of Health under Award Number P20GM104420. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
Developing statistical models and computer simulations to tackle science’s reproducibility crisis
Project Team: Berna Devezer (PI), Erkan Buzbas (Modeler)
Reproducibility of scientific findings has long been considered a pillar of science. However, in the last decade, many disciplines life sciences have failed to reproduce major research findings. This reproducibility crisis has triggered a shift to revise current research practices. Examples include how to make biomedical findings more reproducible as emphasized in a recent article by NIH directors (Collins and Tabak 2014, Policy: NIH plans to enhance reproducibility, Nature 505, 612–613) and the Cancer Biology Reproducibility Project (https://osf.io/e81xl/wiki/home/). Despite these self-correction efforts, little is known about the underpinnings of reproducibility. The goal of this project is to help generate more true research claims than false, by identifying and examining the factors contributing to the non-reproducibility of experimental findings.
Drug-target interactions of veranamine and synthetic analogs with sigma-1 and 5HT2B receptors
Project Team: Jakob Magolan (PI), Xiong Zhang, Jagdish Patel (Modeler)
Depression is a common psychiatric disorder. According to the most prevalent monoamine theory, depression is caused by an imbalance of brain neurotransmitters: serotonin, noradrenaline, and dopamine. All currently available antidepressant drugs act mainly by increasing the concentration of these in the brain. Primary drawbacks of current antidepressant drugs are undesirable side effects and late onset of action. Additionally, these drugs are characterized by relatively low efficacy. There is no ideal medication on the market and new antidepressant drugs are needed. Veranamine is a natural product isolated from the marine sponge Verongida rigida by Hamann and co-workers. Anti-depression activity of veranamine was evaluated using a forced swim test. Veranamine showed potent antidepressant activity at a dose of 20 mg/kg, i.p. decreasing the
immobility time significantly compared to the clinically utilized drug desipramine. A locomotor activity test was also performed to exclude the possibility of nonspecific stimulant action that could create a false-positive read out in the forced swim test. These results indicate potent antidepressant activity that is not a consequence of the compound acting as a stimulant. Veranamine was also tested in a number of receptor binding assays (via the Psychoactive Drug Screening Program; a service provided by the NIMH) revealing affinity specific to 5-HT2B and sigma-1 receptors with Ki values of 388 nM and 557 nM, respectively. This project will synthesize numerous structural analogs of veranamine and evaluate them using receptor binding assays (via the Psychoactive Drug Screening Program provided free of charge by the NIMH). Our goal is to synthesize new compounds based on the veranamine scaffold with increased affinity for 5-HT2B and sigma-1 receptors. The project is also interested in maintaining selectivity for these two receptors and modulating selectivity between them.
The effect of genetic variation on the interaction between primate lentiviruses and host proteins
Project Team: Paul Rowley (PI), Jagdish Patel (Modeler)
The interaction of host proteins with lentiviruses and other retroviruses and retrotransposons represents a major research theme of the Rowley lab. The proposed work fits well with the mission of the NIH and is a field of study that has traditionally been well supported. Dr. Rowley is fully committed to providing all the necessary support for the development of the molecular models and the production of supporting experimental data. Investigating the complex interactions that occur between the capsid protein of primate lentiviruses (including HIV) and human proteins, to understand the consequences of host evolution on viral replication. Then to identify key determinants of lentivirus-primate interaction, specifically between the Nup153p capsid-interacting motif and the HIV-1 capsid. Use known crystal structures as templates for modeling the polymorphisms present within Nup153 variants.
Evolution of tandemly-replicated opsin genes: Molecular models that predict spectral shift
Project Team: Deborah Stenkamp (PI), Robert Mackin (Modeler), Jagdish Patel (Modeler)
Gene replication is an established mechanism for the generation of raw genetic material upon which evolution may act. For example, tandem replication of genes to generate arrays of paralogs underlies functional diversification in vertebrate sensory systems. Tandem replication of opsin (visual pigment) genes and subsequent neofunctionalization provide selective advantages for the exploitation of novel visual environments, food sources, and mate selection. Despite their importance, the mechanisms underlying the subsequent “acts” of neo-functionalization of new genetic material are not clear. In this Modeling Access Grant proposal, we use the tandemly-replicated cone opsin genes of teleosts and primates to address this significant knowledge gap. The tandemly-replicated cone opsin genes are ideal for this study because many independent tandem replications have occurred very recently, and because experimental protein structures are available to inform molecular models. Early in vertebrate evolution, one-rod opsin (RH1) and four cone opsin.
Mathematical modeling of human motions using recurrent neural networks
Project Team: Alex Vakanski (PI), Stephen Lee (Modeler), Jake Ferguson (Modeler)
The current methods in the literature for representing human motions are based on modeling the movements at a single level of abstraction, either at a low-level (i.e., trajectory level) of abstraction or at the high-level (i.e., symbolic level) of abstraction. The proposed project will exploit the latest advances in recurrent neural networks for modeling human motions at multiple hierarchical levels of abstraction. The ultimate aim is to allow patients to perform rehabilitation exercises at home using a sensory system for capturing the motions, where an algorithm will retrieve the trajectories of patient’s exercises, will perform a data analysis by comparing the performed motions to a model of desired motions, and will send the analysis results to the patient’s physician with recommendations for improvement.
Mathematical modeling of nutritional factors and bacterial communities of the maternal-infant dyad
Project Team: Mark McGuire (Project Director), Janet Williams (Graduate Student), Sarah Brooker (Graduate Student), Chris Remien (Collaborator), Ben Ridenhour (Collaborator), JT Van Leuven (Post Doc)
Myriad microbial communities within and on the human body interact constantly with each other and environmental factors, and although much work has focused on characterizing these various community structures (especially in adults), almost nothing is known about how those of mothers and infants interact. Understanding this crosstalk is likely critical to understanding the establishment of the gastrointestinal (GI) microbiome in early life and how it influences risk for diseases and conditions such as necrotizing enterocolitis, diarrhea, obesity, and Crohn’s disease. However, development of GI microbial community in infancy has been relatively poorly studied. Nonetheless, its complexity is known to be substantial and tracking individual species may not provide the knowledge to impact human health.
Here, we propose to model the development of breastfed infants’ GI tract in connection with various sites within the mother-infant dyad and examine potential effect modifiers, such as maternal/infant nutrition, mode/location of delivery, and antibiotic use. For instance, studies have demonstrated that the bacterial community in milk may influence the development of infant’s GI tract. Other studies have shown that maternal consumption of targeted probiotics can alter the bacteria in the milk she produces. Thus, it is possible that the bacterial community in a woman’s GI tract can also alter that of her infant. Other sites of within a woman, such as the skin and saliva, may contribute to milk and infant fecal microbiomes as well. For these reasons, we have collected samples from mothers (milk, breast skin, feces and saliva) and infants (feces and saliva) from birth through 6 mo postpartum to examine these and other relationships. In this project, we will leverage this complex dataset (which includes both quantitative and qualitative empirical data) to model never-before-evaluated relationships among environmental/behavioral factors and multiple microbial communities in mothers and their infants over time. Because of the intimate and emerging connections between microbial communities and health, we anticipate that these findings will lay a solid foundation supporting additional collaborative studies related to the manipulation of early-life microbial communities for optimal acute and chronic health.
Small area estimation of obesity-related indicators
Project Team: Helen Brown (PI), Christopher Murphy (Collaborator), Chantal Vella (Collaborator), Marco Mesa-Frias (Modeler), Michelle Wiest (Modeler)
This project will aim to develop or adapt an existing model to generate small area estimates in Idaho counties of the factors that place individuals at highest risk for obesity (e.g. sedentary behaviors, food insecurity, sugar-sweetened beverage consumption). Knowledge of the local environment is often critical in public health planning and development. In the model, selected demographic characteristics, health conditions, health behaviors, and health status will be estimated to provide a precise geographical picture of the relevant local population. The model will attempt to synthesize a geographically-relevant study population in their local context using various datasets. The American Community Survey (e.g. census) will be used to provide the geographical and demographical context at the local level, and it will be linked to regional datasets such as the Behavioral Risk Factor Surveillance System (BRFFS) to provide domain relevant information, mainly in the context of health-related lifestyles. Prevalence of obesity indicators and other risk factors will be estimated for Idaho counties and local authorities in 2013/2014 to identify the “hot spots” or those counties with the highest risk of obesity
Externally Funded Projects
Collaborative Research: Deep-sequencing analysis of edited metabolic pathways
to uncover, model, and overcome the epistatic constraints upon optimization
Project Team: Christopher J. Marx (University of Idaho), Ankur Dalia (Indiana University), Sergey Stolyar (University of Idaho), Jeremy Draghi (Brooklyn College), Norma Martinez-Gomez (Michigan State University)
Epistasis – non-linear interactions between genotypic changes upon phenotypes –represents a critical challenge to optimization of biological systems, whether by evolution or engineered via synthetic biology. When mutational effects upon growth or product generation depend on the genetic background, assessing performance across the entire parameter space of any system of realistic size quickly becomes impossible. This is especially problematic when there is sign epistasis – mutations that change from beneficial to deleterious depending upon the other loci – as this creates ridges and peaks on the fitness landscape that can restrict stepwise optimization via either synthetic biological changes or beneficial mutations. Development of kinetic computational models of metabolism can provide guidance, but unfortunately these models are dogged by numerous free parameters. There is an immediate need for two linked developments: empirical techniques that can rapidly generate and assess rational, combinatorial variants, and modeling techniques to incorporate these data and predict where in parameter space further rounds of generation and assessment of variation would be most effective. The test-bed for our novel approach is to optimize the function of the high-efficiency ribulose monophosphate (RuMP) pathway that the team has successfully introduced into the model methanolconsuming organism, Methylobacterium extorquens. First, in vivo gene editing of a plasmid-encoded suite of enzymes will be performed, and deep sequencing used to rapidly assess the fitnesses of a quartermillion genotypes with combinatorial variation in nine dimensions of expression. Second, this massive volume of data about epistasis – combined with direct measurement of intracellular metabolite concentrations for select combinations – will be used to infer the numerous parameter values in our kinetic model. Third, the model will be utilized to predict which regions of parameter space would be more or less evolvable and these will be targeted and compared in the second round of editing, fitness assays, and experimental evolution.
NIH R00 Project: Causal Inference of Gene Regulatory Networks with Application to Breast Cancer
Project Team: Audrey Fu (PI), Bahadur Badsha (Post Doc), Rui Li (Post Doc), Evan Martin (Graduate Student)
In the investigation of the mechanisms behind gene regulation and its impact on diseases, two lines of research have been largely separately carried out in recent years. On the one hand, gene regulatory networks and protein interaction networks have been under extensive study, especially in systems biology, where genetic variation is usually ignored. On the other hand, mutations, indels (insertions and deletions), and copy number variants have been identified for many diseases in genome-wide association studies. It is therefore of immense interest to understand how genetic variation influences disease through gene regulatory networks.
To construct these networks, at least three key pieces of information are important: gene expression, transcription factor binding, and genotypes (especially at expression quantitative trait loci; that is, eQTLs). In particular, the latter two enable causal inference in the network construction, although how to use them in a probabilistic and rigorous way has not been systematically explored. This project aims to develop statistical models and efficient computational strategies, drawing on recent advances in graphical models and causal inference, to construct causal regulatory networks involving genetic variation and TF binding. This project will use breast cancer as a disease model and apply the proposed methodologies to different subtypes. Topological features of the inferred regulatory networks may suggest potentially different mechanisms in breast cancer subtypes. Research reported in this publication was supported by the National Human Genome Research Institute of the National Institutes of Health under Award Number R00HG007368. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
Collaborative Research: A Mathematical Theory of Transmissible Vaccines
Project Team: Scott Nuismer (PI), Chris Remien (Co-PI), James Bull (Co-PI), Andrew Basinski (Post Doc)
Viral vaccines have had remarkable and long-lasting impacts on human health, resulting in the worldwide eradication of smallpox, the elimination of polio within much of the developed world, and the effective control of many other diseases. Although great strides have been made in the development and production of vaccines since Edward Jenner’s first vaccinations with cowpox in the early 1800’s, little has changed in the way vaccines are delivered. Even today, virtually every vaccine must be given directly to the patient. Recent advances in molecular biology suggest that the centuries-old method of individual-based vaccine delivery could be on the cusp of a major revolution. Specifically, genetic engineering brings to life the possibility of a “transmissible vaccine.” Rather than directly vaccinating every individual within a population, a transmissible vaccine would allow large swaths of the population to be vaccinated effortlessly by releasing an infectious agent that is genetically engineered to be benign yet infectious. In fact, some existing vaccines are transmissible to a limited extent, and transmissible vaccines have already been developed and deployed in wild animal populations. Remarkably enough, however, no theory exists to guide the safe and effective use of this revolutionary new type of vaccine. We will develop a mathematical framework for understanding the ecology and evolution of transmissible vaccines, and test the emerging mathematical results using an experimental viral system. Epidemiological efficacy will be assessed by calculating the gains in disease protection conferred by a transmissible vaccine relative to a traditional vaccine. Evolutionary robustness will be explored using models that predict the rate at which a genetically engineered vaccine will lose its efficacy or increase its virulence. In both cases, models will be analyzed using a combination of direct and asymptotic solutions, approximations, numerical solutions, and individual-based simulations. Key mathematical results will be tested experimentally using interactions between bacteria and viruses that infect them. Research reported in this publication was supported by the National Institute of General Medical Sciences of the National Institutes of Health under Award Number R01GM122079.