Dataset Information

Quantifying cell densities and biovolumes of phytoplankton communities and functional groups using scanning flow cytometry, machine learning and unsupervised clustering.

ABSTRACT: Scanning flow cytometry (SFCM) is characterized by the measurement of time-resolved pulses of fluorescence and scattering, enabling the high-throughput quantification of phytoplankton morphology and pigmentation. Quantifying variation at the single cell and colony level improves our ability to understand dynamics in natural communities. Automated high-frequency monitoring of these communities is presently limited by the absence of repeatable, rapid protocols to analyse SFCM datasets, where images of individual particles are not available. Here we demonstrate a repeatable, semi-automated method to (1) rapidly clean SFCM data from a phytoplankton community by removing signals that do not belong to live phytoplankton cells, (2) classify individual cells into trait clusters that correspond to functional groups, and (3) quantify the biovolumes of individual cells, the total biovolume of the whole community and the total biovolumes of the major functional groups. Our method involves the development of training datasets using lab cultures, the use of an unsupervised clustering algorithm to identify trait clusters, and machine learning tools (random forests) to (1) evaluate variable importance, (2) classify data points, and (3) estimate biovolumes of individual cells. We provide example datasets and R code for our analytical approach that can be adapted for analysis of datasets from other flow cytometers or scanning flow cytometers.

SUBMITTER: Thomas MK

PROVIDER: S-EPMC5945019 | biostudies-literature | 2018

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Quantifying cell densities and biovolumes of phytoplankton communities and functional groups using scanning flow cytometry, machine learning and unsupervised clustering.

Thomas Mridul K MK Fontana Simone S Reyes Marta M Pomati Francesco F

PloS one 20180510 5

Scanning flow cytometry (SFCM) is characterized by the measurement of time-resolved pulses of fluorescence and scattering, enabling the high-throughput quantification of phytoplankton morphology and pigmentation. Quantifying variation at the single cell and colony level improves our ability to understand dynamics in natural communities. Automated high-frequency monitoring of these communities is presently limited by the absence of repeatable, rapid protocols to analyse SFCM datasets, where image ...[more]

PMID: 29746500

Similar Datasets

Project description:BackgroundThere are many important clustering questions in computational biology for which no satisfactory method exists. Automated clustering algorithms, when applied to large, multidimensional datasets, such as flow cytometry data, prove unsatisfactory in terms of speed, problems with local minima or cluster shape bias. Model-based approaches are restricted by the assumptions of the fitting functions. Furthermore, model based clustering requires serial clustering for all cluster numbers within a user defined interval. The final cluster number is then selected by various criteria. These supervised serial clustering methods are time consuming and frequently different criteria result in different optimal cluster numbers. Various unsupervised heuristic approaches that have been developed such as affinity propagation are too expensive to be applied to datasets on the order of 106 points that are often generated by high throughput experiments.ResultsTo circumvent these limitations, we developed a new, unsupervised density contour clustering algorithm, called Misty Mountain, that is based on percolation theory and that efficiently analyzes large data sets. The approach can be envisioned as a progressive top-down removal of clouds covering a data histogram relief map to identify clusters by the appearance of statistically distinct peaks and ridges. This is a parallel clustering method that finds every cluster after analyzing only once the cross sections of the histogram. The overall run time for the composite steps of the algorithm increases linearly by the number of data points. The clustering of 106 data points in 2D data space takes place within about 15 seconds on a standard laptop PC. Comparison of the performance of this algorithm with other state of the art automated flow cytometry gating methods indicate that Misty Mountain provides substantial improvements in both run time and in the accuracy of cluster assignment.ConclusionsMisty Mountain is fast, unbiased for cluster shape, identifies stable clusters and is robust to noise. It provides a useful, general solution for multidimensional clustering problems. We demonstrate its suitability for automated gating of flow cytometry data.

Project description:In ecology and evolution, the primary challenge in understanding the processes that shape biodiversity is to assess the relationship between the phenotypic traits of organisms and the environment. Here we tested for selection on physio-morphological traits measured by scanning flow-cytometry at the individual level in phytoplankton communities under a temporally changing biotic and abiotic environment. Our aim was to study how high-frequency temporal changes in the environment influence biodiversity dynamics in a natural community. We focused on a spring bloom in Lake Zurich (Switzerland), characterized by rapid changes in phytoplankton, water conditions, nutrients and grazing (mainly mediated by herbivore ciliates). We described bloom dynamics in terms of taxonomic and trait-based diversity and found that diversity dynamics of trait-based groups were more pronounced than those of identified phytoplankton taxa. We characterized the linkage between measured phytoplankton traits, abiotic environmental factors and abundance of the main grazers and observed weak but significant correlations between changing abiotic and biotic conditions and measured size-related and fluorescence-related traits. We tested for deviations in observed community-wide distributions of focal traits from random patterns and found evidence for both clustering and even spacing of traits, occurring sporadically over the time series. Patterns were consistent with environmental filtering and phenotypic divergence under herbivore pressure, respectively. Size-related traits showed significant even spacing during the peak of herbivore abundance, suggesting that morphology-related traits were under selection from grazing. Pigment distribution within cells and colonies appeared instead to be associated with acclimation to temperature and water chemistry. We found support for trade-offs among grazing resistance and environmental tolerance traits, as well as for substantial periods of dynamics in which our measured traits were not under selection.

Project description:Adaptive immune responses are complex dynamic processes whereby B and T cells undergo division and differentiation triggered by pathogenic stimuli. Deregulation of the response can lead to severe consequences for the host organism ranging from immune deficiencies to autoimmunity. Tracking cell division and differentiation by flow cytometry using fluorescent probes is a major method for measuring progression of lymphocyte responses, both in vitro and in vivo. In turn, mathematical modeling of cell numbers derived from such measurements has led to significant biological discoveries, and plays an increasingly important role in lymphocyte research. Fitting an appropriate parameterized model to such data is the goal of these studies but significant challenges are presented by the variability in measurements. This variation results from the sum of experimental noise and intrinsic probabilistic differences in cells and is difficult to characterize analytically. Current model fitting methods adopt different simplifying assumptions to describe the distribution of such measurements and these assumptions have not been tested directly. To help inform the choice and application of appropriate methods of model fitting to such data we studied the errors associated with flow cytometry measurements from a wide variety of experiments. We found that the mean and variance of the noise were related by a power law with an exponent between 1.3 and 1.8 for different datasets. This violated the assumptions inherent to commonly used least squares, linear variance scaling and log-transformation based methods. As a result of these findings we propose a new measurement model that we justify both theoretically, from the maximum entropy standpoint, and empirically using collected data. Our evaluation suggests that the new model can be reliably used for model fitting across a variety of conditions. Our work provides a foundation for modeling measurements in flow cytometry experiments thus facilitating progress in quantitative studies of lymphocyte responses.

Project description:Natural microbial communities affect human life in countless ways, ranging from global biogeochemical cycles to the treatment of wastewater and health via the human microbiome. In order to probe, monitor, and eventually control these communities, fast detection and evaluation methods are required. In order to facilitate rapid community analysis and monitor a community's dynamic behavior with high resolution, we here apply community flow cytometry, which provides single-cell-based high-dimensional data characterizing communities with high acuity over time. To interpret time series data, we draw inspiration from macroecology, in which a rich set of concepts has been developed for describing population dynamics. We focus on the stability paradigm as a promising candidate to interpret such data in an intuitive and actionable way and present a rapid workflow to monitor stability properties of complex microbial ecosystems. Based on single-cell data, we compute the stability properties resistance, resilience, displacement speed, and elasticity. For resilience, we also introduce a method which can be implemented for continuous online community monitoring. The proposed workflow was tested in a long-term continuous reactor experiment employing both an artificial and a complex microbial community, which were exposed to identical short-term disturbances. The computed stability properties uncovered the superior stability of the complex community and demonstrated the global applicability of the protocol to any microbiome. The workflow is able to support high temporal sample densities below bacterial generation times. This may provide new opportunities to unravel unknown ecological paradigms of natural microbial communities, with applications to environmental, biotechnological, and health-related microbiomes. IMPORTANCE Microbial communities drive many processes which affect human well-being directly, as in the human microbiome, or indirectly, as in natural environments or in biotechnological applications. Due to their complexity, their dynamics over time is difficult to monitor, and current sequence-based approaches are limited with respect to the temporal resolution. However, in order to eventually control microbial community dynamics, monitoring schemes of high temporal resolution are required. Flow cytometry provides single-cell-based data in the required temporal resolution, and we here use such data to compute stability properties as easy to interpret univariate indicators of microbial community dynamics. Such monitoring tools will allow for a fast, continuous, and cost-effective screening of stability states of microbiomes. Applicable to various environments, including bioreactors, surface water, and the human body, it will contribute to the development of control schemes to manipulate microbial community structures and performances.

Dataset Information

Quantifying cell densities and biovolumes of phytoplankton communities and functional groups using scanning flow cytometry, machine learning and unsupervised clustering.

Publications

Quantifying cell densities and biovolumes of phytoplankton communities and functional groups using scanning flow cytometry, machine learning and unsupervised clustering.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets