The Finland-based team combine mass spectrometry with computational methods to identify the most abundant proteins in a sample in which microbiota functionality and dynamics can be better defined.
“Until recently, the research on microbiota has strongly focused on discovering which microbes are present in a sample but analysing the functionality of the microbiota has been challenging,” explains Professor Laura Elo, study leader and Research Director at the Turku Bioscience Centre in Finland.
“Recent technological advancements have however also enabled a deeper dive into the functionalities.”
“The study of the protein levels in microbiota samples is one such rising research field, making it possible for us to reach a broader understanding of the functionality and dynamics of microbiota,” adds Professor Elo.
Data-dependent acquisition methods
Writing in the Nature journal ISME Communications, the team discuss the drawbacks of previously used data-dependent acquisition (DDA) mass spectrometry, mainly the limited reproducibility when analysing samples with complex microbial composition.
Along with Postdoctoral Researchers Sami Pietilä and Tomi Suomi, the three team members introduce, for the first time, an untargeted DIA metaproteomics tool that does not require any DDA data.
The technique instead generates a pseudospectral library directly from the DIA data reducing the amount of required mass spectrometry data to a single DIA run per sample.
The team say that the new DIA-only metaproteomics approach will be available as an open-source software package named glaDIAtor, that includes a web-based graphical user interface to encourage the wider use of the tool by the community.
“The new method we have developed for analysing complex protein data produces more reliable results than previous methods,” explains Postdoctoral Researcher Sami Pietilä.
“The currently used research methods typically only analyse the most abundant proteins, which causes fluctuation in the results from one analysis to another.
“The new method analyses the samples systematically and produces reliable results without this type of fluctuation,” Pietilä continues.
In the study’s discussion, the team identify an interesting future development that would involve circumventing the need to generate reference spectra separately for each new project.
Here, machine learning is suggested as a possible solution using, for instance, artificial neural networks.
“A major challenge with such approaches is, however, their potential biases towards the training data and need for re-training for specific conditions,” the team writes. “This remains an interesting topic for further investigation.”
“In general, the microbiome research still involves multiple different types of unknowns that are continuously being revealed thanks to improved technologies.
“Metaproteomics provides an excellent opportunity to uncover the functional aspects of the microbial communities.”
Tomi Suomi adds: “It has been extremely important for us to bring this newly developed method available for all researchers as an easy-to-use application.
“We are also prepared to engage in further development and maintenance of this tool.”
Source: ISME COMMs (Nature)
Published online: doi.org/10.1038/s43705-022-00137-0
“Introducing untargeted data-independent acquisition for metaproteomics of complex microbial samples.”
Authors: Pietilä, S., Suomi, T. & Elo, L.L.