Epidemiological and ecological modelling reveal diversity in upper respiratory tract microbial population structures from a cross-sectional community swabbing study

Abigail L. Coughtrie; Denise E. Morris; Rebecca Anderson; Nelupha Begum; David W. Cleary; Saul N. Faust; Johanna M. Jefferies; Alex R. Kraaijeveld; Michael V. Moore; Mark A. Mullee; Paul J. Roderick; Andrew Tuck; Robert N. Whittaker; Ho Ming Yuen; C. Patrick Doncaster; Stuart C. Clarke

Nestedness is a measure of organisation within an ecological system that quantifies the ordering of species incidence across space or time. A strongly nested system has a highly predictable sequence of species incidence amongst samples, while a weakly nested system has a more random turnover of species incidence. A nestedness tool, PlotTemp.R, was used within the R graphical user interface (CRAN, version 3.0.1) to determine spatial ordering amongst nose-swab samples analysed with real-time PCR. Real-time PCR datasets were converted to incidence matrices of the presence/absence of each species in each sample. The plot_temp script shuffles each matrix to rank its rows of samples by their richness (most species rich at top) and its columns of species by their incidence (most frequently present to left). Within the final shuffled matrix, presences are represented by red squares and absences by white squares. Their locations in the matrix are compared to a line of perfect fill that would encompass all presences if they were perfectly packed towards the top-left of the matrix. This line then reveals the locations of 'surprise' presences lying to its right and 'surprise' absences lying to its left. The degree of disorder in the matrix is reported as a matrix 'temperature', ranging from 0-100°, calculated as a function of the sum across all surprises of squared deviations from the line of perfect fill. Low temperatures reveal strong nesting, characteristic of highly ordered systems, and high temperatures reveal weak nesting, characteristic of disordered systems. Nestedness was compared between groups of individuals partitioned by age, recent RTI and season.