IceR improves proteome coverage and data completeness in global and single-cell proteomics

Abstract

Label-free proteomics by data-dependent acquisition (DDA) enables the unbiased quantification of thousands of proteins, however it notoriously suffers from high rates of missing values, thus prohibiting consistent protein quantification across large sample cohorts. To solve this, we here present IceR, an efficient and user-friendly quantification workflow that combines high identification rates of DDA with low missing value rates similar to DIA. Specifically, IceR uses ion current information in DDA data for a hybrid peptide identification propagation (PIP) approach with superior quantification precision, accuracy, reliability and data completeness compared to other quantitative workflows. We demonstrate greatly improved quantification sensitivity on published plasma and single-cell proteomics data, enhancing the number of reliably quantified proteins, improving discriminability between single-cell populations, and allowing reconstruction of a developmental trajectory. IceR will be useful to improve performance of large scale global as well as low-input proteomics applications, facilitated by its availability as an easy-to-use R-package.

Driving Single Cell Proteomics Forward with Innovation

“…Approaches for peptide sequence identification offer a tradeoff between sensitivity and specificity. Approaches using extracted ion current, such as IceR, 22 are likely to offer higher sensitivity of peptide sequence propagation. In contrast, approaches using more features, such as DART-ID, 21 are likely to offer higher reliability of peptide sequence identification.…”

Replication of single-cell proteomics data reveals important computational challenges

“…Several approaches have been developed to reduce this bias by propagating spectrum identifications from one sample to the corresponding MS1 peaks from another sample. The match between run algorithm of MaxQuant is very popular in label-free SCP [10,9,7,8,34,12], but methodological improvements have recently been suggested for both label-free and TMT-based SCP [35,36]. Finally, another reason for missingness is the inability to match a spectrum to a peptide sequence due to poor spectrum quality.…”

Replication of single-cell proteomics data reveals important computational challenges

“…Several approaches have been developed to reduce this bias by propagating spectrum identifications from one sample to the corresponding MS1 peaks from another sample. The match between run algorithm of MaxQuant is very popular in label-free SCP [10, 9, 7, 8, 34, 12], but methodological improvements have recently been suggested for both label-free and TMT-based SCP [35, 36]. Finally, another reason for missingness is the inability to match a spectrum to a peptide sequence due to poor spectrum quality.…”

“…Several approaches have been developed to reduce this bias by propagating spectrum identications from one sample to MS1 peaks from another sample. The match between run algorithm of MaxQuant (Tyanova et al (2016)) is very popular in label-free SCP (Zhu et al (2018a), Zhu et al (2018b), , Cong et al (2020), Cong et al (2021), Brunner et al (2020), but methodological improvements have recently been suggested for both label-free (Kalxdorf et al (2020)) and TMTbased SCP (Yu et al (2020)). Finally, another reason for missingness is the inability to match a spectrum to a peptide sequence.…”