Improved quantitative accuracy in data-independent acquisition proteomics via retention time boundary imputation

The traditional approaches to handling missing values in data-independent acquisition (DIA) proteomics are to either remove high-missingness proteins or impute missing values with statistical procedures. Both approaches have their disadvantages: removal can limit statistical power, whereas imputation can introduce spurious correlations or dilute signal. We present an alternative approach based on imputing peptide retention times (RTs) rather than quantitations. For each missing value, we impute the start and end of the peptide’s RT profile (“RT boundaries”), then obtain a quantitative value by integrating the chromatographic signal within the imputed boundaries. We evaluate our method on three distinct datasets and show that RT boundary imputation produces more accurate quantitations than traditional imputation methods and reduces peptide lower limit of quantitation. Additionally, RT boundary imputation allows quantitative ratios to be obtained between experimental groups in cases where this would otherwise not be possible, as demonstrated with matrix matched calibration curve and Alzheimer’s disease datasets. Finally, we show that RT boundary imputation improves the ability to estimate radiation exposure in biological tissues. Our RT boundary imputation method, called Nettle, is available as a standalone tool with an open-source software license.