Accounting for multiple imputation-induced variability for differential analysis in mass spectrometry-based label-free quantitative proteomics

Abstract

Imputing missing values is common practice in label-free quantitative proteomics. Imputation aims at replacing a missing value with a user-defined one. However, the imputation itself may not be optimally considered downstream of the imputation process, as imputed datasets are often considered as if they had always been complete. Hence, the uncertainty due to the imputation is not adequately taken into account. We provide a rigorous multiple imputation strategy, leading to a less biased estimation of the parameters’ variability thanks to Rubin’s rules. The imputation-based peptide’s intensities’ variance estimator is then moderated using Bayesian hierarchical models. This estimator is finally included in moderated t-test statistics to provide differential analyses results. This workflow can be used both at peptide and protein-level in quantification datasets. Indeed, an aggregation step is included for protein-level results based on peptide-level quantification data. Our methodology, named mi4p, was compared to the state-of-the-art limma workflow implemented in the DAPAR R package, both on simulated and real datasets. We observed a trade-off between sensitivity and specificity, while the overall performance of mi4p outperforms DAPAR in terms of F-Score.

Proximity Mapping of CCP6 Reveals Its Association with Centrosome Organization and Cilium Assembly

“…Afterward, proteins with less than 2 valid values in at least one group were removed. Assuming the presence of missing data due to the limit of detection, missing values were imputed from a normal distribution around the detection limit (with 0.3 spread and 1.8 down-shift) [ 94 , 95 ]. Afterward, the data were subjected to a scaled median absolute deviation (SMAD) normalization with the R package “marray” (version 1.60.0).…”

Revisiting the Thorny Issue of Missing Values in Single-Cell Proteomics

“…Multiple imputation approaches attempt to tackle this issue by computing multiple predictions for each entry. These predictions are then summarized, typically by averaging, leading to estimates that are more robust than using a single prediction. , Furthermore, the variance associated with each imputation and across imputations can be combined and used for downstream analysis, although this substantially complicates the analytical approach. The performance of multiple imputation however still relies on the performance of the underlying imputation method used and does not solve systematic bias.…”

Augmented Doubly Robust Post-Imputation Inference for Proteomic data

“…There have been some attempts to apply this framework to proteomic data (Yin et al, 2016; Gianetto et al, 2020). Some noticeable challenges involved in using this approach include its empirical conservativeness (Chion et al, 2022), computational burden (Brini and van den Heuvel, 2023), and the lack of a straightforward expression for test statistics (Meng, 1994).…”