WebofMicrobes analyzing the chemistry between microbes


Generation of this data was supported by the DOE Office of Science (Office of Biological and Environmental Research) primarily as part of the ENIGMA SFA (enigma.lbl.gov), Banfield and Firestone carbon cycling programs, Northen LDRD and ECRP, as well as a number of related projects and collaborations.

Electron micrograph of fungal mycelium courtesy Tippkoetter, Rolf & Eickhorst, Thilo. Micropedology – The hidden world of soils. University of Bremen, Germany. http://www.microped.uni-bremen.de

Download the Data

WoM data is available as an SQLite database. We recommend you begin by running the .schema command at the SQLite command line.

About WoM

WebofMicrobes (WoM) is a repository of experimental assertions on how microorganisms transform their environments through the uptake and release of metabolites and exoenzymes. These assertions are based on direct experimental evidence obtained through mass spectrometry based exometabolomic analysis performed by the Northen Lab. WoM captures three types of experimental assertions: metabolite compositions of environments, uptake and release of metabolites by organisms in this environment, and relevant protein activities in this environment (e.g. glycoside hydrolases). It is important to note that these transformations are only interpretable in the context of the specific environments since an organism may uptake a metabolite in one environment but not in another.


The WebofMicrobes is focused on understanding how microorganisms transform their environment. The terms ‘environment’, ‘organism’, and ‘compound’ are used very loosely and therefore require some definition:

‘Environment’ is the starting metabolite pool that is being treated by the ‘organism’ or ‘protein’. Environments may include anything from synthetic mixtures of metabolites, rich medias, plant exudates, microbial extracts, soil extracts, plant materials etc.

‘Organisms’ includes anything except a purified protein that acts upon metabolites in an environment. For example, we have performed experiments where we compare a bacterial extract before and following treatment with soil. In this case the environment is the starting bacterial extract, the ‘organism’ is the soil and the data reflects the change in the metabolite pool resulting from the soil treatment.

‘Compound’ includes any organic materials detectable using our mass spectrometry methods, primarily metabolites that are produced and consumed by microorganisms.


Assertions of presence in an environment: For a metabolite to be annotated as present it must be detected in >2/3 replicates with peak area >10k.

Assertions of uptake or release: To make an assertion the p-value < 0.05 using a 2 tailed equal variance test except for cases where the metabolite is not detected in one group, in which case a one-sample t-test is used. Table elements are colored by the log2 of the fold change in the environment composition (Final/Initial) where zero values are replaced with 1 to prevent undefined results. The darker the color the larger the log2 fold change in that particular metabolite using the system: Score 1<1.5; Score 2= 1.5-2.5; Score 3=2.5-3.5; Score 4=3.5-4.5; Score 5 >4.5

Caveats: Our goal is to make assertions of presence not abundance! Given dramatic differences in instrument response functions for metabolites it is not possible to use these data to compare abundances of different metabolites. Some abundant metabolites are presumably not detected at all because they don’t ionize well (e.g. hydrocarbons). We use the terms ‘takes in metabolite’ and ‘releases metabolite’ to describe our observations that a metabolite was depleted or added to the environment after treatment with the organism. It is important to keep this in mind when interpreting our assertions. For example, if the ‘organism’ is a soil then ‘takes in metabolite’ would include non-biological adsorption onto mineral surfaces.

Using WebofMicrobes

In many cases, the data in the WebofMicrobes has three dimensions: assertions about metabolites in an environment by an organism or community. Therefore it is setup to enable the data to constrain one of these dimensions such that the data can be viewed as two dimensional tables of assertions. Our assertion scoring system is as follows:

Searching: Note that only part of the name of a metabolite, organism etc is required for searching. For example ‘s’ can be used to search for all organisms starting with the letter ‘s’ and searching compounds for ‘aden’ returns adenine, adenosine, methyladenosine, etc.

Metabolite names and IDs: In many cases we are confident making a metabolite identification based on exact mass (+/- 5ppm), retention times vs. authentic standards retention times (+/-30s) and fragmentation spectra (presence of characteristic fragments). In these cases we use the common metabolite name, e.g. adenosine. In other cases, we have some evidence for making an assignment (e.g. partial fragmentation spectra) and we indicate this by putting the name in parentheses, e.g. (adenosine) or otherwise indicate whatever we feel we can say about the metabolite (hexos(amine) trisaccharide). The minimum identifier included in WoM is the chemical formula for a metabolite. Experimental information on the metabolite can be viewed by clicking on the metabolite within the table cell including the protocols used for analysis.


One Environment: This view allows the user to select a single environment and view both the metabolites present in an environment and compare the activities of organisms within this environment. Select an environment and either search for organisms or click on “Select all organisms in this environment” to see the list of all organisms associated with this environment.

Compatibility of organisms with the environment: provides a measure of the compatibility of an organism with the metabolites in an environment based on the number of metabolites in the environment that are acted upon by the microbe. The reference environment is displayed in yellow and is changed by clicking on the column heading. The compatibility of each organism vs. this environment is displayed following the name in the column heading.

(uptaken - released) / total.

Compatibility between organisms in one environment: If more than one organism is selected a simple analysis is automatically performed to compare the compatibility of the organisms with each other in the environment based on the number of correlated vs. anticorrelated metabolites normalized to the total number of uptaken or released metabolites. Here the reference organism is indicated in yellow and can be changed by clicking on it and the numbers displayed are relative to that organism. The score for each of the organisms vs. the reference microbe is displayed after the organisms name. A negative number suggests some competition for metabolites and a positive number suggest that the microbes may be compatible.

(anticorrelated - correlated) / (anticorrelated + correlated)

One Compound: This is the metabolite-centric view of the data that allows the user to view the activities of organisms or proteins associated with a single metabolite with three user defined inputs. First, select either ‘organism’ or ‘protein’. Second search for the compound of interest. Again partial searches can be used, i.e. searching for ‘a’ will return all names containing ‘a’. Third, select the environments of interest. Once this is done WoM will immediately display the activities for all organisms against this metabolite for the selected environments.

One Organism: This view allows the user view all assertions for a particular organism and generate a table of all metabolites across environments of interest. First, select the organism of interest. Second, select the environments and the WoM will generate the table of all assertions.

Upload: Currently only the NorthenLab is uploading data esp. given that relatively few labs are doing these types of experiments. We anticipate that at some point a small network of labs will all contribute data. Please email us if your group is interested in contributing data.

Errors: Please email northenlab@gmail.com if you catch any errors and note that we are unable to provide any support for how to use or interpret WoM assertions.

Modeling efforts: We are working with the DOE KBase (kbase.us) such that data from the WoM repository can be linked to genomics, models etc within the KBase environment.