DAIME: Digital Image Analysis in Microbial Ecology
daime is a scientific image analysis and visualization program for microbiology and microbial ecology. It offers many tools for analyzing 2D and 3D microscopy datasets of microorganisms stained by FISH with rRNA-targeted probes or other fluorescence labelling techniques.
daime has been used in hundreds of studies in microbial ecology, medical microbiology, and environmental engineering. Its visualization features were used to render the cover illustrations of "Brock - Biology of Microorganisms" (12th edition) and of PNAS vol. 103(7).
Free for use in academia
daime is free for use in academic research and education at non-profit institutions. Other users can contact us for non-academic licenses.
How to get daime
daime is available for Windows and Linux operating systems. As a stand-alone program it does not require the installation of other software. Mac users: Please run daime in a virtualized Linux or Windows environment until a native Mac version becomes available.
Selected Program features
- Image segmentation
- Microbial abundance quantification
- FISH probe evaluation
- Microbial co-aggregation pattern analysis
- Virtual biofilm sectioning
- 2D and 3D visualization
Daims H, Lücker S, Wagner M. 2006. daime, a novel image analysis program for microbial ecology and biofilm research. Environ. Microbiol. 8: 200-213. PubMed
daime is written in C++. It is developed and coded by Holger Daims.
This step detects objects (e.g., microbial cells or cell aggregates) in images for subsequent analysis. daime can detect 2D objects in single images and image batches, and 3D objects in z-stacks. It offers different algorithms for fully automated 2D and 3D image segmentation. Manual image segmentation is also possible based on user-defined intensity thresholds or by point-and-click with the mouse at objects in the images.
daime also contains tools for segmenting multicolor images based on object colors (e.g., microbial populations stained by combinations of rRNA-targeted FISH probes).
Once images are segmented, various features of the 2D or 3D objects can be measured. Examples include object size, brightness, surface area, volume, and other important parameters. And the segmented images are the input data for more complex functions of daime such as microbial abundance quantification, FISH probe evaluation, and co-aggregation pattern analyses.
daime contains a powerful object editor that is part of the 2D and 3D visualization module. Here, users can interactively select objects in 2D images or 3D z-stacks for all kinds of analyses. The 3D algorithms allow users to virtually "walk through" a 3D z-stack and pick 3D objects by point-and-click with the mouse.
daime can count objects (cells, cell clusters) in 2D and 3D images. In addition, it can quantify the biovolume fraction of a population, relative to the total biovolume of all microbes, from batches of 2D-segmented images. One image batch shows a specific FISH probe signal, and the other batch shows the signal of the general bacterial probe mix or a suitable nucleic acid stain. This stereological approach estimates a 3D parameter (biovolume fraction) from 2D data (the 2D images). It does not require z-stacks and is accurate if the 2D images are taken at randomly chosen positions in the sample. The biovolume fraction is often more informative than cell numbers, because it represents the "biochemical reaction space" occupied by a microbial population. In other words: Few large cells may have the same biovolume as many small cells.
When a new rRNA-targeted FISH probe is designed, the optimal stringency conditions for specific hybridization must be determined experimentally. Usually this is done in a series of FISH experiments with target and non-target organisms and increasing formamide (FA) concentrations in the hybridization buffers. daime offers a special option to evaluate such FA series. It measures the fluorescence intensities of the cells in a batch of images, which were taken after FISH at the different FA concentrations. From the average intensities the melting curve of the probe is determined. As always in daime, the results can be exported for use in third-party spreadsheet or plotting software.
Specific spatial arrangement patterns of microbes in biofilms (or other highly structured samples) can indicate biological interactions between different microorganisms. For example, symbionts likely co- aggregate, whereas competitors may rather avoid each other.
As spatial arrangement patterns are often complex and subtle, it can be extremely difficult and unreliable to spot such patterns just by visual observation of the samples by microscopy. daime offers a suite of unique stereological algorithms that quantify the spatial arrangement of microbial populations, which have been labeled by specific FISH probes or other fluorescent markers. This works with batches of 2D-segmented images or 3D-segmented confocal z-stacks.
Stratification of biofilms, flocs, or granules often reflects the ecophysiological requirements of the microorganisms in the different depth zones. Not only the abundance, but also spatial arrangement patterns and other features of microbial populations may differ among the layers of a stratified biofilm. To quantify stratification-related phenomena, daime can virtually section biofilm images. Once the images are sliced, all image analysis functions of daime can be applied to characterize the microbial populations in each depth zone.
The image slicing algorithm offers much flexibility, as the direction of slicing, thickness of the produced sections, and several other parameters can be adjusted. Since up to four slicing directions can be combined, even spherical structures can be sectioned correctly.
Confocal microscopy comes to life when z-stacks are projected in 3D and are explored interactively. daime provides everything needed for 3D visualization on commodity PCs and even on most laptops. High-speed volume rendering algorithms display z-stacks and allow free rotation and virtual "fly-through" in real time. Freely adjustable volume rendering parameters ensure that the important features of 3D datasets can be visualized. Hidden structures are exposed by semi-transparent surface rendering or by clipping, and the 3D impression is greatly improved by virtual lighting with realistic shadows. Stereo anaglyphs can be rendered at interactive frame rates and popular features such as maximum intensity projection are included, too.
Multiple z-stacks can easily be combined in the same 3D scene. daime can render beautiful still images, and complex animation sequences can be defined interactively with a convenient user interface. Rendered images and animations can be exported for use in presentations and publications.