Computational protocol: Cytometric fingerprints: evaluation of new tools for analyzing microbial community dynamics

Similar protocols

Protocol publication

[…] CyBar performs a segregated analysis of cytometric histograms without any image analysis step. In this procedure, like in most analyzing procedures in FCM, an operator dependent, and thus experience based gating step has to be performed. Every cluster of cells in a histogram is marked with a gate. The individual gates of each sample are combined to one gate template for a data set. Such a gate template can comprise up to 30 gates and more when natural microbial community data sets are explored (Koch et al., ). The gate template serves then as a mask which is applied to all samples of the data set. The cell abundances in each gate are easily extracted for all samples. Therewith, the abundance variation per gate can directly be compared between samples of different treatments or over a time course. The direct comparison of cell abundance variations between gates with high and low cell numbers is facilitated by data normalization. The dynamic variations of abundances per gate are then visualized in form of a barcode like heat map, the CyBar plot (performed in R, Figure ). The CyBar plot allows identifying stable or highly fluctuating subsets of cells. In this way, a segregated analysis of individual cell cluster responses is possible in addition to the general trend interpretation analysis which was already provided by Dalmatian Plot and CHIC. Moreover, index subcommunities can be identified, i.e., potential functions assigned to clusters of cells by correlation analysis, and sorted for further analysis. A detailed step by step procedure and ready-to-use macros for the CyBar procedure are provided in Koch et al. () and were recently published as R package on the Bioconductor platform ( as flowCyBar, The link is also available under the QR-Code provided in Figure . [...] FlowFP (Rogers and Holyst, ) is a software package of the Bioconductor platform (Gentleman et al., ). Thereby, the complete analyzing procedure can be performed in R. FlowFP was first developed for handling FCM data sets for medical research, but was recently also successfully applied to a microbiological data set (De Roy et al., ). The FlowFP analyzing procedure does not require an image analysis step or any manual gating decision but works on the basis of a geometrical grid. The application uses a probability distribution function to define two regions of the FCM histogram that contain an equal number of cells. These regions are considered as bins and further partitioned with the identical probability distribution function creating equal sub-bins with identical virtual cell numbers. This procedure is repeated for every bin, based on a predefined number of recursions. The result is a geometrical grid with fixed numbers and positions of bins. Consequently, bins in regions with high abundance of virtual cells are smaller compared to those covering regions with low cell abundance. The grid can be built based on one sample or a set of samples. Subsequently, the computed grid serves as a mask which is applied to a complete data set. The number of cells per bin is extracted and stable and fluctuating bins can be identified. Therewith, segregated dynamics within microbial communities can be investigated as well as similarity analyses performed. For extensive information on FlowFP see Rogers and Holyst () and ( A ready-to-use macro for the application of FlowFP to microbial cytometric fingerprints based on FSC and DAPI-DNA fluorescence (application example below) was created and is available under the QR-Code provided in Figure and the following link […]

Pipeline specifications

Software tools flowCyBar, flowFP
Application Flow cytometry
Chemicals Lactic Acid