A predictive model for vertebrate bone identification from collagen using proteomic mass spectrometry.

Area of Science:

• Proteomics
• Bioinformatics
• Forensic Science

Background:

• Proteogenomics enables rapid and cost-effective species identification from biological samples, including degraded ones.
• Tandem mass spectrometry is commonly used to analyze peptides for species identification, relying on species-specific peptides.
• Analyzing bone proteomes presents challenges due to limited protein content and potential for misidentification.

Purpose of the Study:

• To evaluate the causes of peptide sequence errors in proteogenomic analysis of bone samples.
• To develop an unbiased, probabilistic approach for reliable species identification from bone.
• To overcome limitations in current mass spectrometry and algorithmic interpretation for accurate species determination.

Main Methods:

• Analysis of tandem mass spectrometry data from modern human bone samples.
• Investigation of peptide spectra misidentifications against vertebrate databases.
• Development and application of a probabilistic model to assess species identification likelihood.

Main Results:

• A significant percentage of confident spectra from human bone samples were initially misidentified as non-human species.
• Peptide sequence errors are attributed to limitations in mass spectrometry and interpretation algorithms.
• The proposed probabilistic approach provides a more reliable method for species identification without solely relying on species-specific peptides.

Conclusions:

• Current proteogenomic methods can lead to inaccurate species identification from bone due to peptide sequencing errors.
• A probabilistic framework is essential for robustly determining species identity from bone proteomic data.
• This study advances the reliability of species identification in challenging forensic and archaeological contexts.