OutlierD

OutlierD detects outliers in high-throughput mass spectrometry proteomics data using quantile regression on M-A scatterplots to improve preprocessing and downstream analysis.

Key Features:

• Quantile regression on M-A scatterplots: Applies quantile regression to M-A scatterplots for nuanced outlier identification in intensity-dependent comparisons.
• Regression types: Implements linear, non-linear, and non-parametric quantile regression techniques.
• Heterogeneous variability handling: Accommodates heterogeneous variability and both linear and non-linear relationships to reduce false positives in outlier detection.
• Target data: Focused on high-throughput mass spectrometry data used in proteomics analyses.
• Implementation environment: Implemented in R with relevance to the Bioconductor ecosystem.

Scientific Applications:

• Proteomics preprocessing: Identification and removal of outliers during preprocessing of mass spectrometry proteomics datasets.
• Data quality control: Improving the reliability and interpretability of downstream analyses by mitigating variability-driven artifacts.

Methodology:

Quantile regression is applied to M-A scatterplots using linear, non-linear, and non-parametric quantile regression methods; the software is implemented in R.