Progress 09/01/14 to 08/31/19
Outputs
Target Audience:Several portions of the results have been given at international conferences and peer review publications in refereed journals (see Products). The research consortium has organized a symposium and series of workshops that attracted approximately 100 peer scientists. This activity has subsequently resulted in further collaborations between scientists in the field of mycobacterial diseases and infection dynamics.Several portions of the results have been given at international conferences and peer review publications in refereed journals (see Products). The research consortium has organized a symposium and series of workshops that attracted approximately 100 peer scientists. This activity has subsequently resulted in further collaborations between scientists in the field of mycobacterial diseases and infection dynamics. Changes/Problems:
Nothing Reported
What opportunities for training and professional development has the project provided?Training and professional development that has been provided during the project included disease specific knowledge through mentoring by individual faculty members and attending the whole research team face-to-face meetings [NVSL, Ames, IA, March 23-24, 2015; Cornell, Ithaca, NY, June 13-15, 2016; University of Glasgow, Glasgow UK, 4-7 June 2017 (Bovine TB Mini-Symposium at the University of Glasgow, Glasgow UK, 8 June 2017), and Wageningen University, the Netherlands, May 22-24, 2018. (A Symposium and Workshop entitled "Dynamics of mycobacterial infections" in Wageningen University, the Netherlands, May 22-24, 2018. It was very well received. The one-day symposium attracted 64 people and 30 people attended three half-day (1.5 days in total) workshop sessions). In addition, Dr. Grohn has given a 3-day infectious disease modeling workshop in Obihiro, Japan (March, 2017, 2018, 2019) and Oslo, Norway (May 8-10, 2017). Drs. Grohn and Kao attended NSF EEID workshop in London March 18-19, 2019. How have the results been disseminated to communities of interest?Several portions of the results have been given at international conferences and peer review publications in refereed journals (see Products). The research consortium has organized a symposium and series of workshops that attracted approximately 100 peer scientists. This activity has subsequently resulted in further collaborations between scientists in the field of mycobacterial diseases and infection dynamics. What do you plan to do during the next reporting period to accomplish the goals?
Nothing Reported
Impacts
What was accomplished under these goals?
The whole research team had conference calls during the 5th year and the last face-to-face meeting was held at Wageningen University on May 21-25 2018. At the meeting, each team gave update on their on-going research addressing three questions: What has been done? What is planned? What is missing? Cornell's focused on mathematical and economic modeling of Johne's disease, MAP sequencing data and how to use it in connection with detailed epidemiological data. Minnesota reported their research on cattle movement including data from Uruguay and the United States. Glasgow and Michigan explained how to use whole-genome sequencing data to study the role of elk in the bovine Tuberculosis transmission in Michigan, USA. Following the meeting, a 2-day workshop and symposium was held to share the methods developed with the scientific community. The final annual presentation of the five year project was given at the annual Conference of Research Workers in Chicago on November 5, 2019. 1. Modeling Cornell and Illinois In Year 5, previously developed compartment models (Smith et al., 2015, 2016, 2017; Rossi et al., 2017) were used to address whether the effect of MAP infection on clinical mastitis in dairy cows changed the cost-efficacy of control programs (Verteramo Chiu et al., 2019). The herd level agent based model (Al-Mamun et al., 2016; Verteramo Chiu et al., 2018) was adapted to include tuberculosis (Verteramo Chiu et al., 2019) and to explore further the control of paratuberculosis (Al-Mamun et al., 2018). Our previous work did not address the role of the environment in Johne's disease, which is a topic of our current research (Ceres et al., 2017). Johne's disease progression through discrete states was demonstrated using Hidden Markov models (Ceres et al. under review). Minnesota Mycobacterium bovis: We demonstrated limitations of two Phylodynamics software programs, Bayesian Evolutionary Analysis Sampling Trees (BEAST) and SeqTrack that are routinely used for transmission analyses of infectious diseases. We found that these two popular methods developed for viral infections cannot provide reliable inference on groups of closely related M. bovis genomes. For example, in the analysis using SeqTrack, we found that this method is highly sensitive to sampling dates. In the analysis using BEAST, between-species inference is difficult to validate when sampling bias is high among cattle and deer samples. However, for clusters of M. bovis genomes that are distinctive in genetic distances, BEAST gives stable inference even when alternating our hypothesis of the index herd. Mycobacterium avium paratuberculosis: We identified a total of 365 MAP samples that we collected from 2003 to 2009 from 8 herds. One hundred and six samples from 5 herds were sequenced; no results could be obtained from fecal samples. Because of the limitation of the SeqTrack method (it cannot deal with situations when a host is infected with multiple strains), we developing a computational method to identify those mixed infected cows by analyzing the WGS data isolated from MAP genomes. Previous research demonstrated the slow-evolving nature of mycobacterium genomes and suggested that shared single-nucleotide polymorphism(s) (SNP(s)) between animals could be considered as evidence of transmission. Consequently, we believe that sharing multiple evolutionary paths with others suggests that an individual may have been infected by multiple individuals. Using a full dataset of 525 WGS MAP samples, we identified 5 strains and estimated ~10% of the animals were infected with more than 1 strain. We developed a suite of easily executable open-source R functions to transform raw sequence data into actionable information for epidemiological inferences. Various visualizations were provided for quick and easy inspection of the results. Glasgow and Michigan We developed a framework to study bovine tuberculosis transmission in the Lower Peninsula of the state of Michigan. In particular: 1) study the historic and contemporary transmission dynamics of bovine tuberculosis by doing a comparative analysis of genomic pathogen data using whole genome sequencing collected during the last two decades to understand the underlying evolutionary processes of Mycobacterium bovis. 2) investigate bovine tuberculosis transmission in domestic and multiple wildlife hosts using a discrete trait analysis to disentangle who infected whom and estimate rates of pathogen transition between species. 3) perform a phylogeographic analysis of bovine tuberculosis spread to understand the origins and distribution of different pathogen lineages, estimate the rate of the disease spatial spread, and study the processes that may responsible for their contemporary geographic distribution. Combined deer-elk-cattle M. bovis sequence analysis was published in Molecular Ecology (Salvador et al. 2019): central role of deer in maintaining TB in MI over observed period (1999-2013) and that there is a strong association between deer and cattle TB, and between deer and elk, but no observed role for elk in infection of cattle. Sequencing Cornell Various population genetic and genomic analyses have been completed for the 395 sequenced MAP strains arising from the set of three RDQMA farms. For the first time, the core and accessory genome components have been determined for MAP and the functional groups for each of these components (COG: Clusters of Orthologous Groups and KEGG: Kyoto Encyclopedia of Genes and Genomes) have been determined and compared to other groups of Mycobacterium. In addition to these population genomic analyses, a bacteria GWAS (Genome-wide association study) analysis using the program "Scoary" was implemented to correlate cow disease phenotypes to MAP accessory genes. This included a total of 15 different disease related cow phenotypes, derived from the final highly curated version of the RDQMA database. The phenotypes included such aspects as super shedder, disease progression, milk production, MAP infection in tissues etc. A large number of MAP genes were found to be significantly correlated with cow phenotypes, and detailed inspection of the biochemical functions of these genes reveals that many are known virulence factors for other Mycobacterium spp. and/or other bacteria, with specific functions that can sometimes be linked to their physiological and biochemical roles in these host animals. This is a novel analysis on at least two fronts: 1) very few datasets like RDQMA exist for any pathogen host and certainly this is the only one for MAP; 2) bacteria pan-GWAS is new and the connection to very high quality epidemiological data is among the very first projects that will be able to do so; "Scoary" is very new, judged to be state of the art, and to our knowledge it has not been used to correlate host phenotypes to bacteria genes derived from those host individuals. This population genomic + genotype/phenotype analysis will comprise a single publication in a high impact publication. Phylogenetic analysis of the SNP data arising from the RDQMA MAP strains is currently underway for a publication that will address questions regarding paths of disease transmission within the farms. In these publications the connection between detailed epidemiological data and high quality GWAS data will be further explored. We plan to develop models that allow us to answer in this population "Who infects Whom?" Results were presented at the European Veterinary Epidemiology meeting in the spring of 2019. Subsequently the results are currently prepared for submission for publication in a veterinary epidemiology journal. A follow-up publication with results from multiple farms is currently in the process of completion.
Publications
- Type:
Journal Articles
Status:
Published
Year Published:
2019
Citation:
Verteramo Chiu, L.J., Tauer, L.W., Smith, R.L., Gr�hn Y.T.: Assessment of the bovine tuberculosis elimination protocol in the United States. J. Dairy Sci. 2019, 102:1-17. (https://doi.org/10.3168/jds.2018-14990)
- Type:
Journal Articles
Status:
Published
Year Published:
2019
Citation:
14. Verteramo Chiu, L. J., Tauer, L. W., Grohn, Y. T., and Smith, R. L.: Mastitis risk effect on the economic consequences of paratuberculosis control in dairy cattle: A stochastic modeling study. PLOS ONE 2019, 14(9): e0217888. https://doi.org/10.1371/journal.pone.0217888
- Type:
Journal Articles
Status:
Published
Year Published:
2019
Citation:
Salvador L.C.M., OBrien D.J., Cosgrove M.K., Stuber T. P., Schooley A., Crispell J., Church S., Grohn Y.T., Robbe-Austerman S., Kao R.R.: Disease management at the wildlife-livestock interface: using whole-genome sequencing to study the role of elk in Mycobacterium bovis transmission in Michigan, USA. Molecular Ecology 2019; 28: 2192 2205. https://doi.org/10.1111/mec.15061
- Type:
Conference Papers and Presentations
Status:
Published
Year Published:
2019
Citation:
16. Nigsch, A., Robbe-Austerman, S., Stuber, T., Pavinski Bitarc, P.,D., Gr�hn, Y.T., and Ynte H. Schukken. Who infects Whom : Reconstructing infection chains of Johnes disease in an endemically infected dairy herd by use of genomic data. Proceedings SVEPM, Zeist, the Netherlands, 2019.
- Type:
Journal Articles
Status:
Under Review
Year Published:
2019
Citation:
Verteramo Chiu, L.J., Tauer, L.W., Smith, R. L., Gr�hn Y.T.: Economic Effects of a Foodborne Disease: Potential Relationship between MAP in Dairy and Crohns in Humans. Under review.
- Type:
Journal Articles
Status:
Under Review
Year Published:
2019
Citation:
Ceres, K.M., Schukken, Y.H., Gr�hn Y.T.: Characterizing infectious disease progression through discrete states using Hidden Markov models. PlosONE. Under review.
- Type:
Journal Articles
Status:
Under Review
Year Published:
2019
Citation:
Verteramo Chiu, L.J., Tauer, L.W., Gr�hn Y.T., Smith, R.L.: Ranking of epidemiological control strategies with stochastic outcomes.
- Type:
Journal Articles
Status:
Other
Year Published:
2019
Citation:
Richards, V., P., Nigsch, A., Pavinski Bitar, P., Stuber, T., Ceres, K., Robbe Austerman, S., Schukken, Y.H., Gr�hn Y.T., and
Stanhope, M., J.: Evolutionary genomic and bacteria GWAS analysis of Mycobacterium avium subsp. paratuberculosis and dairy cattle Johnes disease phenotypes.
- Type:
Conference Papers and Presentations
Status:
Published
Year Published:
2018
Citation:
13. Verteramo Chiu, L. J., Tauer, L. W., Grohn, Y. T., Smith, R. L.: The Effect of Co-Morbidity on Disease Control Decisions: A Stochastic Modeling Approach. International Symposium on Veterinary Epidemiology and Economics, Chiang Mai, Thailand, November 12-16, 2018.
- Type:
Conference Papers and Presentations
Status:
Published
Year Published:
2018
Citation:
14. Ceres, K., Al-Mamuna, M.A., Gr�hn, Y.T.: Environmental transmission of Mycobacterium Avium Subsp. Paratuberculosis: an individual based model. International Society for Veterinary Epidemiology and Economics Conference, Chiang Mai, Thailand, November 2018
- Type:
Conference Papers and Presentations
Status:
Published
Year Published:
2018
Citation:
15. L.C.M. Salvador, D.J. OBrien, M.K. Cosgrove, T. P. Stuber, A. Schooley, S. Church, Y.T., Grohn, S. Robbe-Austerman, R.R. Kao: Understanding the emergence and evolution of Mycobacterium bovis in Michigan, USA. International Symposium on Veterinary Epidemiology and Economics, Chiang Mai, Thailand, November 12-16, 2018.
- Type:
Conference Papers and Presentations
Status:
Published
Year Published:
2018
Citation:
16. Y.T. Grohn, M. Stanhope, L. Tauer, S. Wells, Y.H. Schukken, and R. Kao. US-UK Collab: Mycobacterial Transmission Dynamics in Agricultural Systems: Integrating Phylogenetics, Epidemiology, Ecology, and Economics USDA NIFA PI's annual meeting, Chicago, December 3-5, 2018.
- Type:
Conference Papers and Presentations
Status:
Published
Year Published:
2019
Citation:
17. Nigsch, A., Robbe-Austerman, S., Stuber, T., Pavinski Bitar, P., Gr�hn Y.T., Schukken, Y.H.: Who infects whom - reconstruction of transmission trees in an endemic phase of mycobacterium avium subsp. paratuberculosis (MAP) infection in dairy herds by use of sequence data. SVEPM 2019 conference paper.
- Type:
Conference Papers and Presentations
Status:
Published
Year Published:
2019
Citation:
18. Y.T. Grohn, M. Stanhope, L. Tauer, S. Wells, Y.H. Schukken, and R. Kao.: US-UK Collab: Mycobacterial Transmission Dynamics in Agricultural Systems: Integrating Phylogenetics, Epidemiology, Ecology, and Economics USDA NIFA PI's annual meeting, Chicago, November 2-5, 2019.
- Type:
Conference Papers and Presentations
Status:
Published
Year Published:
2019
Citation:
Ceres, K, and Gr�hn, Y.T.: Characterizing infection trajectories of slowly progressing infectious disease using hidden Markov models. CRWAD, Chicago, IL, November 2-5, 2019.
- Type:
Conference Papers and Presentations
Status:
Published
Year Published:
2019
Citation:
Y. Wang, M. J. Stanhope , P. Pavinski Bitar, A. Nigsch, Y. H. Schukken , S. J. Wells. Reproducible R functions to analyze WGS data from animals infected with multiple strains of Johnes disease. CWRAD, Chicago, IL, November 2-5, 2019.
- Type:
Conference Papers and Presentations
Status:
Awaiting Publication
Year Published:
2019
Citation:
Salvador L.C.M., OBrien D.J., Cosgrove M.K., Stuber T. P., Schooley A., Church S., Grohn Y.T., Robbe-Austerman S., Kao R.R.: Understanding the emergence and evolution of Mycobacterium bovis in Michigan, USA. Epidemics7 - International Conference on Infectious Disease Dynamics, Charleston, SC, USA, December 3-6, 2019
|
Progress 09/01/17 to 08/31/18
Outputs
Target Audience:Several portions of the results have been given at international conferences and peer review publications in refereed journals (see Products). The research consortium has organized a symposium and series of workshops that attracted approximately 100 peer scientists. This activity has subsequently resulted in further collaborations between scientists in the field of mycobacterial diseases and infection dynamics. Changes/Problems:
Nothing Reported
What opportunities for training and professional development has the project provided?Training and professional development that has been provided during this year of the project included disease specific knowledge through mentoring by individual faculty members and attending the whole research team face-to-face meeting at SCENE, University of Glasgow, Glasgow UK, 4-7 June 2017. A Bovine Tuberculosis Mini-Symposium was organized at the University of Glasgow, Glasgow UK, 8 June 2017. A three-day infectious disease modeling workshop was given in Obihiro, Japan (March 21-24, 2017) and in Oslo, Norway (May 8-10, 2017). A Symposium and Workshop entitled "Dynamics of mycobacterial infections" was organized in Wageningen University, the Netherlands, May 22-24, 2018. It was very well received. The one-day symposium attracted 64 people and 30 people attended three half-day (1.5 days in total) workshop sessions. An independently funded student (Assel Akhmetova, University of Glasgow) has been working with Kao and Salvador on data from Northern Ireland that extends the work being done in this project. How have the results been disseminated to communities of interest?Several portions of the results have been given at international conferences and peer review publications in refereed journals (see Products). The research consortium has organized a symposium and series of workshops that attracted approximately 100 peer scientists. This activity has subsequently resulted in further collaborations between scientists in the field of mycobacterial diseases and infection dynamics. What do you plan to do during the next reporting period to accomplish the goals?Data analysis including sequencing work continues in year 5 and model building activity will gradually increase when the data become available. Our plan is to repeat the 2018 workshop as it was given with some modifications in Edinburgh, 2019 (subject to funding).
Impacts
What was accomplished under these goals?
The whole research team had conference calls during the year and a face-to-face meeting at Wageningen University on May 21-25 2018. At the meeting, each team gave update on their on-going research addressing three questions: What has been done? What is planned? What is missing? Cornell's focused on mathematical and economic modeling of Johne's disease, MAP sequencing data and how to use it in connection with detailed epidemiological data. Minnesota reported their research on cattle movement including data from Uruguay and the United States. Glasgow and Michigan explained how to use whole-genome sequencing data to study the role of elk in the bovine Tuberculosis transmission in Michigan, USA. Following this meeting, a 2-day workshop and symposium was held to share the methods developed with the scientific community. 1. Modeling Cornell and Illinois In Year 4, previously developed compartment models (Smith et al., 2015, 2016, 2017; Rossi et al., 2017) were used to address whether the effect of MAP infection on clinical mastitis in dairy cows changed the cost-efficacy of control programs (Smith et al., in submission). The herd level agent based model (Al-Mamun et al., 2016; Verteramo Chiu et al., 2018) was adapted to include tuberculosis (Verteramo Chiu et al., in submission) and to explore further the control of paratuberculosis (Al-Mamun et al., 2018). Our previous work did not address the role of the environment, which is a topic of our current research (Ceres et al., 2017). Whole genomic sequence data are now beginning to become available and the models will be modified accordingly. Minnesota Mycobacterium bovis: In our analysis of WGS data of the Minnesota bovine TB outbreak (2005 -2012), using BEAST software suites, we estimated the phylodynamics (lower-bound estimates of substitution rate, time to the most recent common ancestor, and the effective population size). Using distance-based phylogenetic trees and network modeling techniques, we observed evidence of a single introduction of bovine TB to the cattle population among infected farms, and cross-species transmission between cattle and white-tail deer. Based this evidence, we conclude that bovine TB in the deer population of MN was a spill-over infection from the cattle population due to shared habitats between species. Mycobacterium avium paratuberculosis: For our development of the computational method to quantify samples with multiple strains from WGS data, the method we have developed uses techniques of multiple sequence alignment to derive ancestor strains from the MAP database, then applies non-negative least squares to calculate the number of strains and mixture coefficients of each strain presented in a given sample. Using synthetic sequencing data incorporating sequencing errors and missing values, we showed that our method has an average accuracy of 84% when the ancestor strains are informative. Glasgow and Michigan The goal of the current study is to develop a framework to study bovine tuberculosis transmission in the Lower Peninsula of the state of Michigan. In particular: 1) study the historic and contemporary transmission dynamics of bovine tuberculosis by doing a comparative analysis of genomic pathogen data using whole genome sequencing collected during the last two decades to understand the underlying evolutionary processes of Mycobacterium bovis. 2) investigate bovine tuberculosis transmission in domestic and multiple wildlife hosts using a discrete trait analysis to disentangle who infected whom and estimate rates of pathogen transition between species. 3) perform a phylogeographic analysis of bovine tuberculosis spread to understand the origins and distribution of different pathogen lineages, estimate the rate of the disease spatial spread, and study the processes that may responsible for their contemporary geographic distribution. A manuscript analyzing a combined deer-elk-cattle M. bovis sequence dataset was submitted to the journal Molecular Ecology and reviewed favorably. A revised manuscript was resubmitted (Salvador et al.). This work identified likely rates of transition between each species pair, and showed evidence for strong interactions between deer and cattle, and deer and elk, suggesting at best a minimal role for elk in the recurrent infections cattle in the high risk areas of Michigan. This work is also being presented at the International Society for Veterinary Epidemiology and Economics or ISVEE conference (Nov 12-16th, 2018). We are currently analyzing a comprehensive dataset that includes 548 cervids, 224 farmed animals, and 87 furbearers from 1994-2017, which will form the basis of a simulation-based phylodynamic approach (modified from related models for bTB phylodynamics in badgers and cattle in GB - O'Hare et al., in prep, and an extant model of TB in Michigan deer - Ramsey et al. 2017). Sequencing Cornell Various population genetic and genomic analyses have now been completed for the 395 sequenced MAP strains arising from the set of three RDQMA farms. For the first time, the core and accessory genome components have been determined for MAP and the functional groups for each of these components (COG: Clusters of Orthologous Groups and KEGG: Kyoto Encyclopedia of Genes and Genomes) have been determined and compared to other groups of Mycobacterium. In addition to these population genomic analyses, a bacteria GWAS (Genome-wide association study) analysis using the program "Scoary" was implemented to correlate cow disease phenotypes to MAP accessory genes. This included a total of 15 different disease related cow phenotypes, derived from the final highly curated version of the RDQMA database. The phenotypes included such aspects as super shedder, disease progression, milk production, MAP infection in tissues etc. A large number of MAP genes were found to be significantly correlated with cow phenotypes, and detailed inspection of the biochemical functions of these genes reveals that many are known virulence factors for other Mycobacterium spp. and/or other bacteria, with specific functions that can sometimes be linked to their physiological and biochemical roles in these host animals. This is a novel analysis on at least two fronts: 1) very few datasets like RDQMA exist for any pathogen host and certainly this is the only one for MAP; 2) bacteria pan-GWAS is new and the connection to very high quality epidemiological data is among the very first projects that will be able to do so; "Scoary" is very new, judged to be state of the art, and to our knowledge it has not been used to correlate host phenotypes to bacteria genes derived from those host individuals. This population genomic + genotype/phenotype analysis will comprise a single publication, currently in preparation, with some hope of a high impact publication. Phylogenetic analysis of the SNP data arising from the RDQMA MAP strains is currently underway for a second and third publication that will address questions regarding paths of disease transmission within the farms. In these publications the connection between detailed epidemiological data and high quality GWAS data will be further explored. We plan to develop models that allows us to answer in this population Whom infects Whom? A presentation of the results has been accepted for the European Veterinary Epidemiology meeting in the spring of 2019. Subsequently the results will be submitted for a focus issue in a veterinary epidemiology journal.
Publications
- Type:
Journal Articles
Status:
Published
Year Published:
2017
Citation:
Al-Mamun, M.A., Smith, R.L., Schukken, Y.H., Gr�hn, Y.T.: Use of an Individual-based Model to Control Transmission Pathways of Mycobacterium avium Subsp. paratuberculosis Infection in Cattle Herds. Scientific reports. 2017; 7(1):11845. DOI: 10.1038/s41598-017-12078-z
- Type:
Journal Articles
Status:
Published
Year Published:
2018
Citation:
Verteramo Chiu, L.J., Tauer, L.W., Al-Mamun, M. A., Kaniyamattam, K., Smith, R.L., Grohn, Y.T.: An Agent-Based Model Evaluation of Economic Control Strategies for Paratuberculosis in a Dairy Herd. J. Dairy Sci., 2018, 101(7):6443-6454. DOI: 10.3168/jds.2017-13175
- Type:
Journal Articles
Status:
Awaiting Publication
Year Published:
2018
Citation:
Al-Mamun, M.A., Smith, R.L., and Schukken, Y.H., Grohn, Y.T.: A data-driven individual-based model of infectious disease in livestock operation: A validation study for Paratuberculosis in PloS One (https://doi.org/10.1101/394569).
- Type:
Journal Articles
Status:
Under Review
Year Published:
2018
Citation:
Verteramo Chiu, L.J., Tauer, L.W., Gr�hn Y.T., Smith, R.L.: Assessment of the Bovine Tuberculosis Eradication Protocol in the U.S. J. Dairy Sci. (Under review)
- Type:
Journal Articles
Status:
Under Review
Year Published:
2018
Citation:
Verteramo Chiu, L.J., Tauer, L.W., Gr�hn Y.T., Smith, R.L.: Mastitis risk changes the economic consequences of paratuberculosis control in dairy cattle: A stochastic modeling study. Pre. Vet. Med. (Under review)
- Type:
Journal Articles
Status:
Submitted
Year Published:
2018
Citation:
Verteramo Chiu, L.J., Tauer, L.W., Smith, R.L., Gr�hn Y.T.: Economic Effects of a Foodborne Disease: Potential Relationship between MAP in Dairy and Crohns in Humans. Food Policy
- Type:
Conference Papers and Presentations
Status:
Published
Year Published:
2017
Citation:
Wang, Y. Wells, S., (2017, December) Identifying within-host multiple strain infection of Paratuberculosis using Whole genome sequencing data to inform transmission inference. The 98th Annual Conference of Research Workers in Animal Diseases, Chicago, IL, USA.
- Type:
Conference Papers and Presentations
Status:
Published
Year Published:
2017
Citation:
Verteramo Chiu, L. J., Tauer, L. W., Al-Mamun, M. A., Kaniyamattam, K., Smith, R. L., & Grohn, Y. T. An Agent-Based Model Evaluation of Economic Control Strategies for Paratuberculosis in a Dairy Herd. In 2017 Agricultural & Applied Economics Association Annual Meeting. Chicago, IL.
- Type:
Conference Papers and Presentations
Status:
Published
Year Published:
2018
Citation:
Wang, Y. Wells, S., (2018, May) The inferences on strain proportions from mixture samples of Mycobacterial diseases to improve transmission inference. Principles and dynamics governing transmission of mycobacterial infection, Symposium and Workshop, Wageningen University, The Netherlands.
- Type:
Conference Papers and Presentations
Status:
Published
Year Published:
2018
Citation:
Wang, Y. Wells, S. (2018, June) Improving transmission inference for paratuberculosis in multi-strain contexts. The 14th International Colloquium on Paratuberculosis, Riviera Maya, Mexico.
- Type:
Conference Papers and Presentations
Status:
Published
Year Published:
2018
Citation:
Wang, Y. Wells, S., (2018, April) Are cattle infected with multiple strains of MAP? A new computational method to detect from sequencing data. Minnesota Dairy Health Conference, St. Paul, MN, USA.
|
Progress 09/01/16 to 08/31/17
Outputs
Target Audience:Training and professional development that has been provided during this year of the project included disease specific knowledge through mentoring by individual faculty members and attending the whole research team face-to-face meeting at SCENE, University of Glasgow, Glasgow UK, 4-7 June 2017. A Bovine Tuberculosis Mini-Symposium was organized at the University of Glasgow, Glasgow UK, 8 June 2017. A three-day infectious disease modeling workshop was given in Obihiro, Japan (March 21-24, 2017) and in Oslo, Norway (May 8-10, 2017). Changes/Problems:
Nothing Reported
What opportunities for training and professional development has the project provided?Training and professional development that has been provided during this year of the project included disease specific knowledge through mentoring by individual faculty members and attending the whole research team face-to-face meeting at SCENE, University of Glasgow, Glasgow UK, 4-7 June 2017. A Bovine Tuberculosis Mini-Symposium was organized at the University of Glasgow, Glasgow UK, 8 June 2017. A three-day infectious disease modeling workshop was given in Obihiro, Japan (March 21-24, 2017) and in Oslo, Norway (May 8-10, 2017). How have the results been disseminated to communities of interest?Several portions of the results have been given at international conferences and peer review publications in refereed journals (see Products). What do you plan to do during the next reporting period to accomplish the goals?Data analysis including sequencing work continues in year 3 and model building activity will gradually increase when the data become available. Our plan is to organize as a joint-effort a three-day modeling workshop "Dynamics of mycobacterial infections" in the Netherlands, in 2018.
Impacts
What was accomplished under these goals?
The goal of this project is to develop a quantitative methodology for incorporating WGS data into bacterial transmission models for infectious diseases incorporating ecology, economics, molecular biology, and epidemiology. We will apply these methods and models towards better understanding the principles and dynamics governing transmission of mycobacterial infection. We are positioned to do this due to our valuable collections of data and isolates in the US and UK and the unique composition of our team of experts across multiple fields of science. We have 4 specific objectives. First, we will produce a methodology for incorporating WGS-based phylogenetic data into bacterial transmission models. Second, we will produce a methodology for developing models of bacterial disease in systems where parameter estimation will be greatly aided by WGS data. We will use these methodologies to test the general hypothesis that a pathogen's phylogenetic structure is directly informed by its epidemiology, providing a 'signature' that can identify the role of wildlife and environmental reservoirs in transmission. This approach will answer 3 specific hypotheses: 1) wildlife and cattle movement play distinct roles in maintaining bovine tuberculosis in the US and UK, 2) MAP transmission in dairy herds is more complex than originally believed, including crucial contributions from the environment, and 3) farm economics and cost-benefit based decision making affect transmission dynamics and infection control in agricultural systems. Third, we will develop research methods for multi-scale models of infection dynamics. Fourth, we will train postdoctoral researchers, graduate, and undergraduate students in the fields of epidemiology, ecology, economics, and phylogenetics. The whole research team had conference calls during the year and the face-to-face meeting at Glasgow UK, 4-7 June 2017. At the meeting, each team gave update on their on-going research addressing three questions: What has been done? What is planned? What is missing? 1. Modeling Cornell and Illinois In Year 2, developed compartment models (Smith et al., 2015, 2016) were used to address specific objectives: economic consequences of Johne's disease (Smith et al., 2017) and the effect of MAP infection on clinical mastitis in dairy cows (Rossi et al., 2017). Economics was also implemented into a herd level agent based model (Al-Mamun et al., 2016; Verteramo Chiu et al., 2017). Our previous work did not address the role of the environment, which is a topic of our current research (Ceres et al., 2017). Whole genomic sequence data are now beginning to become available and the models will be modified accordingly. We will continue expanding our MAP modeling research to include Bovine TB, as well. Minnesota Mycobacterium bovis: First, we demonstrated limitations of two Phylodynamics software programs, Bayesian Evolutionary Analysis Sampling Trees (BEAST) and SeqTrack that are routinely used for transmission analyses of infectious diseases. We found that these two popular methods developed for viral infections cannot provide reliable inference on groups of closely related M. bovis genomes. For example, in the analysis using SeqTrack, we found that this method is highly sensitive to sampling dates. In the analysis using BEAST, between-species inference is difficult to validate when sampling bias is high among cattle and deer samples. However, for clusters of M. bovis genomes that are distinctive in genetic distances, BEAST gives stable inference even when alternating our hypothesis of the index herd. Mycobacterium avium paratuberculosis: To continue the University of Minnesota's work on within-herd modeling, we identified a total of 365 MAP samples that we collected from 2003 to 2009 from 8 herds in Minnesota for Whole-genome sequencing: 320 isolate samples and 45 fecal samples. In summary, a total of 71 isolates have been sequenced (approximately 22%), 48 isolates are waiting for sequencing, and no results could be obtained from fecal samples. Because of the limitation of the SeqTrack method (it cannot deal with situations when a host is infected with multiple strains), we are developing a computational method to identify those mixed infected cows by analyzing the WGS data isolated from MAP genomes. Previous research demonstrated the slow-evolving nature of mycobacterium genomes and suggested that shared single-nucleotide polymorphism(s) (SNP(s)) between animals could be considered as evidence of transmission. Consequently, we believe that sharing multiple evolutionary paths with others suggests that an individual may have been infected by multiple individuals. Built upon a previous method that analyzes consensus SNPs and allele frequencies, our method uses a combination of distance-based phylogenetic tree and Machine Learning techniques for text mining and clustering gene expression, to calculate mixture coefficients for strain concentration. Glasgow and Michigan The goal of the current study is to develop a framework to study bovine tuberculosis transmission in the Lower Peninsula of the state of Michigan. In particular: 1) study the historic and contemporary transmission dynamics of bovine tuberculosis by doing a comparative analysis of genomic pathogen data using whole genome sequencing collected during the last two decades to understand the underlying evolutionary processes of Mycobacterium bovis. 2) investigate bovine tuberculosis transmission in domestic and multiple wildlife hosts using a discrete trait analysis to disentangle who infected whom and estimate rates of pathogen transition between species. 3) perform a phylogeographic analysis of bovine tuberculosis spread to understand the origins and distribution of different pathogen lineages, estimate the rate of the disease spatial spread, and study the processes that may responsible for their contemporary geographic distribution. Sequencing Cornell The Cornell MAP group has completed all the MAP genome sequencing that will be conducted for this project, including both the three-farm Regional Dairy Quality Management Alliance (RDQMA) dataset (farms from VT, NY and PA) and the Minnesota multi-farm set of samples. The final curated genome sequence data set for RDQMA includes 395 strains. Various analyses of this set of strains have now been completed, with additional analyses ongoing and with the intention that the results will be combined in a single or possibly two publications. The RDQMA dataset of bovine phenotypes has now been updated and all of a wide range of phenotypes for each cow have been entered into a spreadsheet to accompany the MAP strains sampled and sequenced from each cow. MAP genome sequence analyses have determined the MAP core genome and its accessory genome, which appears to be about 11% of a typical genome size. Analyses are underway that will correlate genes from this accessory genome with the bovine phenotypes of the RDQMA database. Examples of other analyses completed on this set of 395 strains include recombination, which revealed a remarkable complete absence of recombination in the core genome, and various population genetic analyses which support a highly significant genetic distinction between the three populations both based on analysis of the core genome and of the accessory genome. Other ongoing analyses include network and phylogenetic analyses designed to assess paths of MAP transmission within herds, in an attempt to address questions related to "who infects whom". MAP sequencing for Minnesota isolates resulted in a set of 119 genome sequences. Analyses of these data are being conducted by the Minnesota group. Culturing and sequencing of MAP from fecal samples was unsuccessful; pure, sequence-able strains could not be cultured.
Publications
- Type:
Journal Articles
Status:
Published
Year Published:
2017
Citation:
Smith, R.L., Al-Mamun, M.A., Gr�hn, Y.T.: Economic consequences of paratuberculosis control in dairy cattle: a stochastic modeling study Prev. Vet. Med. 2017, 138: 17-27.
- Type:
Journal Articles
Status:
Published
Year Published:
2017
Citation:
Rossi, G., Gr�hn, Y.T., Schukken, Y.H., Smith, R.L.: The effect of Mycobacterium avium ssp. paratuberculosis infection on clinical mastitis occurrence in dairy cows. J. Dairy Sci. 2017, 100: 7446-7454.
- Type:
Conference Papers and Presentations
Status:
Accepted
Year Published:
2017
Citation:
Ceres, K., Al-Mamun, M., Grohn, Y. T.: Use of an individual based model to describe the role of environmental transmission of Mycobacterium avium subsp. Paratuberculosis. The Biological & Biomedical Sciences Symposium, Cornell, Ithaca, NY, Aug 18, 2017.
- Type:
Journal Articles
Status:
Under Review
Year Published:
2017
Citation:
4. Salvador, L.C.M., O Brien, D.J., Cosgrove, M.K., Stubber, T.P., Shooley, A., Crispell, J., Church, S., Grohn, Y.T., Robbe-Austerman, S., Kao, R.R.: Implications for disease management at the wildlife-livestock interface: using whole-genome sequencing to study the role of elk in the bovine Tuberculosis transmission in Michigan, USA.
- Type:
Journal Articles
Status:
Under Review
Year Published:
2017
Citation:
Al-Mamun, M.A., Smith, R.L., Schukken, Y.H., Gr�hn, Y.T.: Use of an Individual-based Model to Control Transmission Pathways of Mycobacterium Avium Subsp. Paratuberculosis Infection in Cattle Herds.
- Type:
Journal Articles
Status:
Under Review
Year Published:
2017
Citation:
Verteramo Chiu, L.J., Tauer, L.W., Al-Mamun, M. A., Kaniyamattam, K., Smith, R.L., Grohn, Y. T.: An Agent-Based Model Evaluation of Economic Control Strategies for Paratuberculosis in a Dairy Herd.
- Type:
Conference Papers and Presentations
Status:
Accepted
Year Published:
2016
Citation:
Gr�hn, Y. T., Stanhope, M., Tauer, L., Wells, S., Schukken, Y. H., Kao, R. US-UK Collab: Mycobacterial Transmission Dynamics in Agricultural Systems: Integrating Phylogenetics, Epidemiology, Ecology, and Economics USDA NIFA PI's annual meeting, Chicago, December 2, 2016.
- Type:
Conference Papers and Presentations
Status:
Accepted
Year Published:
2017
Citation:
Al-Mamun, M.A., Gr�hn, Y.T.: MABS: A multiscale agent based model of a dairy herd. Proceeding paper. SpringSim 2017, April 23-26, Virginia Beach, VA, USA; Society for Modeling and Simulation (SCS) International.
- Type:
Conference Papers and Presentations
Status:
Accepted
Year Published:
2017
Citation:
Verteramo Chiu, L. J., Tauer, L. W., Al-Mamun, M. A., Kaniyamattam, K., Smith, R. L., Grohn, Y. T. An Agent-Based Model Evaluation of Economic Control Strategies for Paratuberculosis in a Dairy Herd. In 2017 Agricultural & Applied Economics Association Annual Meeting. Chicago, IL.
|
Progress 09/01/15 to 08/31/16
Outputs
Target Audience:
Nothing Reported
Changes/Problems:
Nothing Reported
What opportunities for training and professional development has the project provided?Training and professional development that has been provided during this year of the project included disease specific knowledge through mentoring by individual faculty members and attending the whole research team face-to-face meeting at Cornell, Ithaca, NY. The Next-Generation-Sequencing and Bovine Tuberculosis Workshop, was organized in Glasgow, UK, 3-4 September 2015. A Cornell Post doc visited Dr. Smith's lab for a week and a Glasgow post doc visited for a month at the Department of Ecology and Evolutionary Biology, Princeton University. How have the results been disseminated to communities of interest?Several portions of the results have been given at international conferences and three publications have been published (see Products). What do you plan to do during the next reporting period to accomplish the goals?Data analysis including sequencing work continues in year 3 and model building activity will gradually increase when the data become available.
Impacts
What was accomplished under these goals?
The whole research team had very successful conference calls and a face-to-face meeting at Cornell, June 13-15, 2016. In the first year of the project the emphasis was on data collection and analysis for both MAP and bovine tuberculosis (bTB) data. At the meeting, we started by making sure that all participants had a good understanding of our ultimate modeling goal. The second day was dedicated to whole-genomic sequencing analysis and the third day focus was on "Putting sequencing data into models for use in economic decision making". 1. Modeling Cornell and Illinois In addition to two research papers on compartment models (Smith et al., 2015, 2016), a herd level agent based model was constructed (Al-Mamun et al., 2016). In June 2016, a post-doc was hired to implement economics into this agent based model. This work was scheduled to begin after data had been collected to incorporate into this model, and those data are now beginning to become available in the project. Once the economics components have been implemented and tested, the model will be used to evaluate various MAP control mechanisms, including various test and cull strategies, maintaining a closed herd, and vaccination. The information on agent to agent transmission will aid in improving the assessment accuracy and allow a more complete understanding of halting the transmission of infection between animals under various controls. Minnesota Mycobacterium bovis: Phylodynamic inferences rely on strong, well-resolved phylogenetic trees providing reliable information on ancestral relationships and divergence time. To evaluate the performance of five commonly used tree-building methods (neighbor-joining, Unweighted Pair Group Method with Arithmetic Mean, maximum parsimony, maximum likelihood, and Bayesian Evolutionary Analysis Sampling Trees), we studied population structures (temporal change of population size) from 86 closely related whole-genome-sequenced isolates of M. bovis collected from 2001 to 2013. Then we simulated the evaluation of M. bovis sequences under nine demographical scenarios. Finally, for each scenario, we constructed phylogenetic trees and evaluated the accuracy and bias of the inferred topology and branch length with the grand truth. We found that: 1) branch length estimation is more challenging than topology reconstruction in closely related sequences; and 2) estimation of tree branches with variable and uncorrelated substitution rates in the two host scenarios is least accurate. Our results suggest that phylogenetic inference alone cannot provide confidence estimation on the evolution of M. bovis or explain the between-herd transmission events of bovine tuberculosis. We expect that identification of mixed infections and integration of epidemiological data will provide more robust results. Mycobacterium avium paratuberculosis: To continue our work on within-herd modeling, we identified 545 MAP samples for Whole genome sequencing, collected from 2003 to 2009 from eight herds in Minnesota: 360 isolate samples and 185 fecal samples. The WGS was completed for 71 isolate samples collected from herds Q and S between 2005 and 2009. Bowtie2 was used for reference mapping to the corrected MAP K10 genome. 1060 variant sites were called using SAMtools on 68 samples that passed quality check using FastQC. Concatenated variant sequences were used to build phylogenetic trees using the bootstrap Maximum likelihood method. We observed a single branch switch between the two herds in 2007. Glasgow and Michigan The goal of the study is to develop a framework to understand the cross-species transmission patterns of bacterial diseases. We have started by analyzing data from M. bovis in cattle and other wildlife species in the state of Michigan. Our major findings are the following: 1) BTB has been circulating in Michigan at least since the 1960's and a minimum of four lineages have been circulating since then, with at least 4 different introductions of the disease in deer, three in elk and two in cattle; 2) The mean evolutionary rate of M. bovis was estimated to be 0.41 substitutions per site per year, consistent with previous M. bovis studies; 3) The phylogeny topology together with an ancestral host-state reconstruction analysis show that there is some clustering of the disease in elk and cattle, indicating the possibility of elk to elk transmission, cattle to cattle transmission or exposure to a common source each; and 4) There is no evidence for significant spreading between elk and cattle, suggesting that cross-species transmission in Michigan is most likely maintained by deer. Sequencing Cornell Isolates received and cultured Over the course of this year we received 165 MAP isolates suitable for subculture (isolates that were too dehydrated or overgrown with fungus were discarded and are not included in this summary), bringing the total to 974. To date we have grown or attempted to grow all 974 samples and have had a success rate of just over 60%, or 600 samples. Sequencing results We sequenced an additional 359 MAP samples, for a total of 510 genome sequences. The genomic DNA was extracted using a modified Epicentre Gram Positive DNA extraction protocol outlined in last year's report. Genome libraries were prepared using the Illumina Nextera XT library preparation kit and sequenced in pools of 48 using the HiSeq 2500 Rapid Run with 100bp paired end reads to an average of about 100-120x coverage. Bioinformatics To provide consistent bioinformatics support across the entire project the Whole-Genome-Sequencing bioinformatics pipeline used to analyze M. bovis and MAP sequence data developed at the USDA's National Veterinary Services Laboratory (NVSL) was reviewed and modified to identify mixed strain infections of MAP and M. bovis, This allowed us to improve the resolution of SNP calling in the pipeline, and also provide valuable epidemiological insight through analysis of samples identified as being mixed strain. In addition to the fecal samples, which are yet to be cultured, we will be sequencing approximately 100 more MN MAP samples that have been growing for the past few months. This should bring the total number of sequences to a minimum 600. SNP calling and population genetic and phylogenetic analyses A set of 414 genome sequences from isolates from NY, PA and VT (NY=172; PA=58; VT=184) were assembled using SPAdes and population genetic and phylogenetic analyses were conducted. Gubbins was used to call SNPs and to assess levels of recombination; the latter was found to be minimal - a total of 128 nucleotides were removed as putative recombinants before phylogenetic reconstruction, resulting in a final alignment of approximately 5000 SNPs. An analysis of molecular variance (AMOVA) and Fisher's Exact test were conducted and revealed highly significant population differentiation between the three populations. Phylogenetic analysis corroborated this and suggested three highly distinct populations with only limited overlap between the three farms. Because of discrepancies between SNP counts arising from the approach adopted for this analysis and that undertaken by our USDA collaborators, and for the sake of consistency across the analyses of this project, we have incorporated the USDA's pipeline in calling the SNPs of the NY, PA, and VT sequences. Phylogenetic analysis will be undertaken shortly. An evaluation of the nature of the discrepancy and of the best technical approach is currently underway. Upon resolution, the phylogenetic analyses will be used in concert with the metadata from the RDQMA database to address questions pertinent to the model development regarding MAP transmission. The Michigan Dept. of Community Health Mycobacteriology Lab revived 240 archived isolates (with 131 left to do), including the recently approved 2014-2015 deer, and shipped them to the USDA's NVSL for whole genome sequencing.
Publications
- Type:
Journal Articles
Status:
Published
Year Published:
2015
Citation:
Smith, R.L., Schukken, Y.H., Gr�hn, Y.T.: A new compartmental model of Mycobacterium avium subsp. paratuberculosis infection dynamics in cattle. Prev. Vet. Med. 2015 Dec 1;122(3):298-305. doi: 10.1016/j.prevetmed.2015.10.008. Epub 2015 Oct 21.
- Type:
Journal Articles
Status:
Published
Year Published:
2016
Citation:
2. Smith, R.L., Gr�hn, Y.T., Pradhan, A.K., Whitlock, R.H., Van Kessel, J.S., Smith, J.M., Wolfgang, D.R., Schukken, Y.H.: The effects of progressing and non-progressing Mycobacterium avium ssp. paratuberculosis infection on milk production in dairy cows. Journal of Dairy Science. 2016, 99: 1383-1390.
- Type:
Journal Articles
Status:
Published
Year Published:
2016
Citation:
3. Al-Mamun, M.A., Smith, R.L., Schukken, Y.H., Gr�hn, Y.T.: Modeling of Mycobacterium Avium Subsp. Paratuberculosis Dynamics in a Dairy Herd: An Individual Based Approach. Journal of Theoretical Biology. 408, pp.105-117.
- Type:
Conference Papers and Presentations
Status:
Published
Year Published:
2015
Citation:
1. Smith, R.L., Schukken, Y.H, Lu, Z., Gr�hn, Y.T.: Rethinking Mycobacterium avium subsp. paratuberculosis infection dynamics in cattle: a new modeling approach. Symposium on Veterinary Epidemiology and Economics, Merida, Yucatan, Mexico, November 2-7, 2015.
- Type:
Conference Papers and Presentations
Status:
Published
Year Published:
2015
Citation:
2. Smith, R.L., Schukken, Y.H, Lu, Z., Gr�hn, Y.T.: Revisiting the effect of Johnes disease status on milk production in dairy cows. Symposium on Veterinary Epidemiology and Economics, Merida, Yucatan, Mexico, November 2-7, 2015.
- Type:
Conference Papers and Presentations
Status:
Published
Year Published:
2015
Citation:
3. Salvador, L.C.M., OBrien, D.J., Cosgrove, M.K., Robbe-Austerman, S., Stuber, T.P., Kao, R.R. Estimating bacterial cross-species transmission in Michigan. Symposium on Veterinary Epidemiology and Economics, Merida, Yucatan, Mexico, November 2-7, 2015.
- Type:
Conference Papers and Presentations
Status:
Published
Year Published:
2015
Citation:
4. Salvador, L.C.M., Deason, M., Bessell, P.R., Enright, J., Kao, R.R. How much bovine Tuberculosis surveillance is needed? A quantitative study in low incidence areas in England. (Symposium on Veterinary Epidemiology and Economics, Merida, Yucatan, Mexico, November 2-7, 2015.
- Type:
Conference Papers and Presentations
Status:
Published
Year Published:
2015
Citation:
5. Y.T. Gr�hn, M. Stanhope, L. Tauer, S. Wells, Y.H. Schukken, and R. Kao. US-UK Collab: Mycobacterial Transmission Dynamics in Agricultural Systems: Integrating Phylogenetics, Epidemiology, Ecology, and Economics USDA NIFA PI's annual meeting, Chicago, December 4, 2015.
- Type:
Conference Papers and Presentations
Status:
Published
Year Published:
2016
Citation:
6. Al-Mamun, M.A., Gr�hn, Y.T.: An Individual Based Model for Mycobacterium Avium Subsp. Paratuberculosis Transmission Dynamics in a Dairy Herd. The 14th annual Ecology and Evolution of Infectious Disease Meeting. June 3-5, 2016, Cornell University.
- Type:
Conference Papers and Presentations
Status:
Published
Year Published:
2016
Citation:
7. Salvador, L.C.M., OBrien, D.J., Cosgrove, M.K., Robbe-Austerman, S., Stuber, T.P., Kao, R.R. The use of Whole-Genome-Sequencing to understand bovine Tuberculosis cross-species transmission patterns in Michigan, USA. The 14th annual Ecology and Evolution of Infectious Disease Meeting. June 3-5, 2016, Cornell University.
|
Progress 09/01/14 to 08/31/15
Outputs
Target Audience:
Nothing Reported
Changes/Problems:
Nothing Reported
What opportunities for training and professional development has the project provided?Training and professional development that has been provided during this year of the project included disease specific knowledge through mentoring by individual faculty members and attending the whole research team face-to-face meeting at the NVSL, Ames, Iowa. How have the results been disseminated to communities of interest?Several portions of the results have been accepted for presentation at the International Symposium for Veterinary Epidemiology and Economics in November 2015. In addition, a number of publications based on this research are currently under review. What do you plan to do during the next reporting period to accomplish the goals?Data analysis including sequencing work continues in year 2 and model building activity will gradually increase when the data become available.
Impacts
What was accomplished under these goals?
The goal of this project is to develop a quantitative methodology for incorporating whole genome sequence (WGS) data into bacterial transmission models for infectious diseases incorporating ecology, economics, molecular biology, and epidemiology. We will apply these methods and models towards better understanding the principles and dynamics governing transmission of mycobacterial infection. We are in position to do this due to our valuable collections of data and isolates in the United States (US) and United Kingdom (UK) and due to the unique composition of our team of experts across multiple fields of science. Our specific objectives fall into 4 categories. First, we will produce a methodology for incorporating WGS-based phylogenetic data into bacterial transmission models. Secondly, we will produce a methodology for developing models of bacterial disease in systems where parameter estimation will be greatly aided by WGS data. We will use these methodologies to test the general hypothesis that a pathogen's phylogenetic structure is directly informed by its epidemiology, providing a 'signature' that can identify the role of wildlife and environmental reservoirs in transmission. This approach will answer 3 specific hypotheses: 1) wildlife and cattle movement play distinct roles in maintaining bovine tuberculosis in the US and UK, 2) MAP transmission in dairy herds is more complex than originally believed, including crucial contributions from the environment, and 3) farm economics and cost-benefit based decision making affect transmission dynamics and infection control in agricultural systems. Thirdly, we will develop research methods for multi-scale models of infection dynamics. Fourthly, we will train postdoctoral researchers, graduate, and undergraduate students in the fields of epidemiology, ecology, economics, and phylogenetics. In addition to conference calls the whole research team met at the NVSL March 23-24, 2015 to discuss the workload for the project. In the first year of the project the emphasis has been on data collection and analysis in both MAP and bTB projects. Bioinformatics To provide consistent bioinformatics support across the entire project the Whole-Genome-Sequencing bioinformatics pipeline used to analyze Mycobacterium bovis and Mycobacterium avium paratuberculosis (MAP) sequence data developed at NVSL was reviewed and modified to identify mixed strain infections of MAP and M. bovis, This objective allowed us to improve the resolution of SNP calling in the pipeline, and also provide valuable epidemiological insight through analysis of samples identified as being mixed strain. Sequencing The Michigan Dept. of Community Health Mycobacteriology Lab, revived 160 archived M. bovis isolates were shipped to USDA's National Veterinary Services Laboratory (NVSL) for whole genome sequencing. Another 21 isolates had already been sequenced at NVSL from duplicate submissions. Together, these 181 isolates (176 deer, 5 elk) constitute 46% of the 394 surviving isolates archived at MDCH. At NVSL, a total of 510 genomes from Michigan have been have sequenced (the breakdown 310 wild deer, 7 game farm deer, 1 farmed bison, 160 cattle, 6 domestic cats, 1 bobcat, 14 raccoons, 11 opossums). At Cornell, good progress has been made on growing and sequencing Mycobacterium avium ssp. paratuberculosis (MAP) isolates from 3 herds from Cornell and 8 herds from Minnesota. To date we have a total of 812 samples of MAP collected over eight years from cows, goats and the environment on six farms in four states. Following bacteria culture and extraction, DNA samples are prepped for next generation sequencing using the Nextera XT DNA library preparation kit by Illumina and sequenced using the HiSEQ 2500 Rapid Run with paired end 100bp reads. To date, we have completed sequencing of 151 strains. Modeling An individual based model (IBM) of MAP infection dynamics of a dairy herd has been in development for the last 7 months. The general framework has been already developed and named as Individual based model for MAP dynamics in dairy herd. The main objective is to see whether MAP dynamics can be modeled using an individual based approach where the individual animal behavior combined with different key processes contributes to the infection context. Currently, the model is in a testing phase where the optimum sets of parameters will be defined to maintain a real herd dynamics in our IBM model. In Glasgow, the model to estimate bacterial cross-species transmission using Michigan data is being developed. The objective of this work is to better understand how the disease is maintained in WTD and how it is transmitted across species. Using epidemiological and Next Generation Sequencing data of Mycobacterium bovis isolates, patterns of disease transmission with genetic similarity (based on SNP differences) between elk, deer and cattle are compared. In Minnesota, to improve the performance of phylogenetic tree building methods, MN received the sequencing results of a total of 84 M. bovis isolates from the USDA APHIS National Veterinary Services Lab, performed the multiple sequence alignment on the short reads based on reference genome of M. bovis and selected 389 SNPs with consistently good alignment results. The 389 SNPs are then concatenated together into new sequences for phylogenetic analysis. To evaluate the three most popular tree building methods (distance based, maximum likelihood and Bayesian) of phylogenetics, they implemented the two distance-based methods (Neighbor Joining and Unweighted Pair Group Method with Arithmetic Mean) in R and obtained the results from two programs (PhyML 3.0 and BEAST software). In Illinois, two studies have been conducted related to understanding Johne's Disease impacts and control. First, the effect of MAP test status on milk production in cows that did and did not progress to high shedding status during the RDQMA study was examined. This research (publication under review) found that the milk production of progressing cows continues to decline after the first positive test, but the milk production of non-progressing cows is negatively affected primarily at the time of the positive test, and will improve as time continues. Second, a mathematical model for MAP was built that separates progressing and non-progressing cows and allows for adult infection. This model was parameterized by RDQMA data, finding that progressing animals spend less time in latency than non-progressing animals. The results of the model (publication under review) show that both adult infection and separate progression pathways can play an important role in the dynamics and persistence of MAP in dairy herds.
Publications
|
|