Dates: August 2011 – June 2014
Project Title: Investigating the effects of movement strategies on the population dynamics of multi-trophic ecological communities
Daniel Ryan (Ph.D. Mathematics, Univ. of Miami, 2011) investigated the role of non-random dispersal strategies in ecological systems and the effects of these dispersal strategies on species distribution and community composition. Ryan used various types of models, including partial differential equations, integro-differential equations (for kernel based dispersal schemes), and individual based models. The models were used to investigate questions about evolutionary stability of movement strategies, stability of coexistence states in multi-trophic communities, and the effects of inhomogeneous environments on species distributions.
Additional work completed during the NIMBioS postdoctoral fellowship included two computationally-focused side projects. The first project developed an algorithm for the in silico design of a specialized library of gene variants that systematically vary codon usage patterns by synonymous substitution. The resulting library of genes should be useful to experimentally test hypotheses about the effects of codon usage bias on gene expression. The library is designed so that it can be assembled in a simultaneous, combinatorial fashion via Gibson Isothermal Assembly. This results in huge savings compared to assembling each gene in the library separately.
The second project used a neural network model to predict mortality risk for intensive care unit (ICU) patients in a hospital. The model takes information collected about the patients in the first 48 hours for their stay in the ICU and uses this to predict a risk of mortality for the remainder of their stay in the hospital. Accurate risk assessment for incoming patients can help to improve both quality of care and improve evaluation of treatment effectiveness.
Following his postdoctoral fellowship, Ryan took a position as a database administrator.
NIMBioS Seminar: Non-random dispersal in intraguild predation communities
Feature Story: Measures of movement: Postdoc explores behavior of movement
Measures of movement
Publications while at NIMBioS
Ryan D, Cantrell RS. 2015. Avoidance behavior in intraguild predation communities: A cross-diffusion model. Discrete and Continuous Dynamical Systems, 35(4): 1641-1663.
Cantrell RS, Cosner C, Lou Y, Ryan D. 2012. Evolutionary stability of ideal free dispersal strategies: A nonlocal dispersal model. Canadian Applied Mathematics Quarterly, 20(2012): 15-38. [Online]
Federico P, Gross LJ, Lenhart S, Ryan D. 2013. Optimal control in individual-based models: Implications from aggregated methods. The American Naturalist, 181(1): 64-77. [Online]
Ryan D, Cantrell RS. Avoidance Behavior in Intraguild Predation Communities: A Cross-diffusion Model. Submitted.
Ryan D, Daley BJ, Wong K, Zhao X. 2013. Prediction of ICU in-hospital mortality using a deep Boltzmann machine and dropout neural net. Proceedings of the Biomedical Sciences and Engineering Conference (BSEC). [Online]
Ryan D, Papamichail D. 2013. Rational design of orthogonal libraries of protein coding genes. ACS Synthetic Biology, 2(5): 237-244. [Online]
Presentations while at NIMBioS
Ryan D. 22 May 2014. Deriving Discrete Adjoint Equations for the Numerical Solution of Optimal Control Problems with Higher Order Integrators. NIMBioS Seminar, University of Tennessee, Knoxville
Ryan D. 23 May 2013. Prediction of ICU In-hospital Mortality Using a Deep Boltzmann Machine and Dropout Neural Net, ORNL Biomedical Sciences and Engineering Conference
Ryan D. 1 March 2013. An Introduction to Deep Learning Biomedical Engineering Student Seminar on Machine Learning, University of Tennessee, Knoxville
Ryan D. 26 July 2012. A Cross-diffusion Model for Avoidance Behavior in an Intraguild Predation Community, Annual Meeting of the Society for Mathematical Biology, Knoxville, Tennessee
Ryan D. 23 July 2012. Numerical Solution of Optimal Control Problems, NIMBioS Seminar, University of Tennessee, Knoxville
Ryan D. 3 July 2012. A Cross-diffusion Model for Avoidance Behavior in an Intraguild Predation Community, 9th AIMS Conference on Dynamical Systems, Differential Equations and Applications, Orlando, FL
Ryan D. 11 October 2011. Nonrandom Dispersal in Intraguild Predation Communities, NIMBioS Seminar University of Tennessee, Knoxville
Ryan D. 12 September 2011. Modeling Avoidance Strategies for Intraguild Prey – Cross-diffusion PDE Systems, Differential Equations and Computational Math Seminar, University of Tennessee, Knoxville
Collaborations, Working Groups, and Workshops
Participant. NIMBioS Working Group: Pretty Darn Good Control
Participant. NIMBioS Investigative Workshop: Disturbance Regimes and Climate-Carbon Feedback
Participant. Challenges in Representation Learning Workshop, Atlanta, GA, 2013
Participant. Machine Learning for System Identification Workshop, Atlanta, GA, 2013
Participant. Deep Learning for Audio, Speech, and Language Processing Workshop, Atlanta, GA, 2013
Participant. Everything Disperses to Miami Workshop, Miami, FL, 2012
Participant. Workshop 5: Spatial Models of Micro and Macro Systems, Mathematical Biosciences Institute Workshop, Columbus, OH, 2012
Established collaborations with Xiaopeng Zhao (UTK Biomedical Engineering) and Calistus Ngonghala (former NIMBioS postdoc), both of which are ongoing
Education, Outreach and Training
Mentored students during the 2012 NIMBioS REU program
Assisted with labs during the 2013 NIMBioS-MBI-CAMBAM graduate summer program
Presented a talk for undergraduates at Fisk University: A Survey of Mathematical Models in Spatial Ecology, Mathematics Dept. Seminar, March 29, 2012
Attended NIMBioS Postdoc Career Development Seminars, NIMBioS Interdisciplinary Seminars, UTK Math Department Seminars and CISML Seminars at UTK.
Participated in informal NIMBioS course on agent-based models