msImpute

msImpute imputes peptide intensity in label-free mass spectrometry proteomics experiments and provides diagnostics for missing data mechanisms and post-imputation distributional distortions.

Key Features:

• MAR Imputation: Missing values are imputed by low-rank approximation of the underlying data matrix when missingness is missing at random (method = "v2").
• MNAR Imputation: Missing values are imputed by the Barycenter approach when missingness is missing not at random (method = "v2-mnar").
• Peptide Identity Propagation: Missing values are imputed by Peptide Identity Propagation (PIP).
• MAR/MNAR Diagnosis: The package contains tools for diagnosing whether missing values are missing at random or missing not at random.
• Post-Imputation Assessment: The package assesses distortions to the probability distribution of the data after imputation.

Scientific Applications:

• Label-Free Mass Spectrometry Proteomics: Imputation of peptide intensity data from label-free mass spectrometry experiments.
• Missing Data Mechanism Characterization: Diagnosis of missing at random versus missing not at random patterns in proteomics datasets.
• Data Quality Control: Assessment of distributional distortions introduced by imputation methods.

Methodology:

Low-rank approximation of the data matrix for MAR data, Barycenter approach for MNAR data, and Peptide Identity Propagation (PIP).