Analysis of bacterial biotyping datasets with a mass-based phylonumerics approach

Area of Science:

• Microbiology
• Bioinformatics
• Mass Spectrometry

Background:

• Biotyping is crucial for bacterial identification.
• Current biotyping platforms often rely on specialized software and hardware.
• Mass spectrometry generates complex proteomic datasets.

Purpose of the Study:

• To implement and evaluate the MassTree algorithm for bacterial biotyping using expressed protein mass data.
• To demonstrate MassTree's capability in constructing phylogenetic trees mirroring established biotyping outputs.
• To present MassTree as an accessible alternative to commercial biotyping platforms.

Main Methods:

• A phylonumerics approach utilizing the MassTree algorithm was applied.
• Expressed protein mass data from two diverse bacterial datasets were used.
• The algorithm was adapted for higher mass inputs and larger mass error tolerances.

Main Results:

• The MassTree algorithm successfully generated phylogenetic trees from mass spectrometry data.
• The resulting trees were comparable to those produced by standard main spectral profile (MSP) dendrograms.
• The approach demonstrated effectiveness across different bacterial datasets.

Conclusions:

• The MassTree algorithm provides a viable, single-step method for bacterial phylogenetic analysis using mass spectrometry data.
• This approach offers a cost-effective and versatile alternative to commercial biotyping solutions.
• MassTree enables the use of data from any mass spectrometer, reducing reliance on proprietary software.