QCQuan: A Web Tool for the Automated Assessment of Protein Expression and Data Quality of Labeled Mass Spectrometry Experiments

In
the context of omics disciplines and especially proteomics and
biomarker discovery, the analysis of a clinical sample using label-based
tandem mass spectrometry (MS) can be affected by sample preparation
effects or by the measurement process itself, resulting in an incorrect
outcome. Detection and correction of these mistakes using state-of-the-art
methods based on mixed models can use large amounts of (computing)
time. MS-based proteomics laboratories are high-throughput and need
to avoid a bottleneck in their quantitative pipeline by quickly discriminating
between high- and low-quality data. To this end we developed an easy-to-use
web-tool called QCQuan (available at qcquan.net) which is built around
the CONSTANd normalization algorithm. It automatically provides the
user with exploratory and quality control information as well as a
differential expression analysis based on conservative, simple statistics.
In this document we describe in detail the scientifically relevant
steps that constitute the workflow
and assess its qualitative and quantitative performance on three reference
data sets. We find that QCQuan provides clear and accurate indications
about the scientific value of both a high- and a low-quality data
set. Moreover, it performed quantitatively better on a third data
set than a comparable workflow assembled using established, reliable
software.