Improved sequence variant analysis strategy by automated false positive removal

Area of Science:

• Biotechnology
• Proteomics
• Bioinformatics

Background:

• Sequence variant analysis (SVA) is crucial for therapeutic protein development, ensuring clone integrity and minimizing cell culture misincorporations.
• Current methods for verifying sequence variants from mass spectrometry data are time-consuming, often requiring weeks of manual effort to filter false positives.

Purpose of the Study:

• To develop and validate an automated method for reducing false positives in sequence variant analysis from mass spectrometry data.
• To improve the efficiency of therapeutic protein characterization by minimizing manual data verification.

Main Methods:

• Development of a Perl-based script to evaluate sequence variant identification hits.
• Application of orthogonal criteria to filter false positives from PepFinder™ (MassAnalyzer) software output.
• Comparison of automated verification results with manual verification processes.

Main Results:

• The Perl script reduced false positives in PepFinder™ output by approximately 4-fold.
• Accuracy in detecting true positive identifications was maintained.
• Manual verification time was reduced by over 70%.

Conclusions:

• Automated filtering of sequence variant analysis data using a Perl script significantly enhances efficiency and reduces false positives.
• This approach streamlines the development of therapeutic proteins by accelerating clone verification.
• The developed script offers a substantial time-saving alternative to manual data review.