Numbers of Exchangeable Hydrogens from LC-MS Data of Heavy Water Metabolically Labeled Samples.

Labeling with deuterium oxide (D2O) has emerged as one of the preferred approaches for measuring the synthesis of individual proteins in vivo. In these experiments, the synthesis rates of proteins are determined by modeling mass shifts in peptides during the labeling period. This modeling depends on a theoretical maximum enrichment determined by the number of labeling sites (NEH) of each amino acid in the peptide sequence. Currently, NEH is determined from one set of published values. However, it has been demonstrated that NEH can differ between species and potentially tissues. The goal of this work was to determine the number of NEH for each amino acid within a given experiment to capture the conditions unique to that experiment. We used four methods to compute the NEH values. To test these approaches, we used two publicly available data sets. In a de novo approach, we compute NEH values and the label enrichment from the abundances of three mass isotopomers. The other three methods use the complete isotope profiles and body water enrichment in deuterium as an input parameter. They determine the NEH values by (1) minimizing the residual sum of squares, (2) from the mole percent excess of labeling, and (3) the time course profile of the depletion of the relative isotope abundance of monoisotope. In the test samples, the method using residual sum of squares performed the best. The methods are implemented in a tool for determining the NEH for each amino acid within a given experiment to use in the determination of protein synthesis rates using D2O.

Exact Integral Formulas for False Discovery Rate and the Variance of False Discovery Proportion.

Multiple hypothesis testing is an integral component of data analysis for large-scale technologies such as proteomics, transcriptomics, or metabolomics, for which the false discovery rate (FDR) and positive FDR (pFDR) have been accepted as error estimation and control measures. The pFDR is the expectation of false discovery proportion (FDP), which refers to the ratio of the number of null hypotheses to that of all rejected hypotheses. In practice, the expectation of ratio is approximated by the ratio of expectation; however, the conditions for transforming the former into the latter have not been investigated. This work derives exact integral expressions for the expectation (pFDR) and variance of FDP. The widely used approximation (ratio of expectations) is shown to be a particular case (in the limit of a large sample size) of the integral formula for pFDR. A recurrence formula is provided to compute the pFDR for a predefined number of null hypotheses. The variance of FDP was approximated for a practical application in peptide identification using forward and reversed protein sequences. The simulations demonstrate that the integral expression exhibits better accuracy than the approximate formula in the case of a small number of hypotheses. For large sample sizes, the pFDRs obtained by the integral expression and approximation do not differ substantially. Applications to proteomics data sets are included.

Flexible Quality Control for Protein Turnover Rates Using d2ome.

Bioinformatics tools are used to estimate in vivo protein turnover rates from the LC-MS data of heavy water labeled samples in high throughput. The quantification includes peak detection and integration in the LC-MS domain of complex input data of the mammalian proteome, which requires the integration of results from different experiments. The existing software tools for the estimation of turnover rate use predefined, built-in, stringent filtering criteria to select well-fitted peptides and determine turnover rates for proteins. The flexible control of filtering and quality measures will help to reduce the effects of fluctuations and interferences to the signals from target peptides while retaining an adequate number of peptides. This work describes an approach for flexible error control and filtering measures implemented in the computational tool d2ome for automating protein turnover rates. The error control measures (based on spectral properties and signal features) reduced the standard deviation and tightened the confidence intervals of the estimated turnover rates.

A large-scale LC-MS dataset of murine liver proteome from time course of heavy water metabolic labeling.

Metabolic stable isotope labeling with heavy water followed by liquid chromatography coupled with mass spectrometry (LC-MS) is a powerful tool for in vivo protein turnover studies. Several algorithms and tools have been developed to determine the turnover rates of peptides and proteins from time-course stable isotope labeling experiments. The availability of benchmark mass spectrometry data is crucial to compare and validate the effectiveness of newly developed techniques and algorithms. In this work, we report a heavy water-labeled LC-MS dataset from the murine liver for protein turnover rate analysis. The dataset contains eighteen mass spectral data with their corresponding database search results from nine different labeling durations and quantification outputs from d2ome+ software. The dataset also contains eight mass spectral data from two-dimensional fractionation experiments on unlabeled samples.

Quantifying label enrichment from two mass isotopomers increases proteome coverage for in vivo protein turnover using heavy water metabolic labeling.

Heavy water metabolic labeling followed by liquid chromatography coupled with mass spectrometry is a powerful high throughput technique for measuring the turnover rates of individual proteins in vivo. The turnover rate is obtained from the exponential decay modeling of the depletion of the monoisotopic relative isotope abundance. We provide theoretical formulas for the time course dynamics of six mass isotopomers and use the formulas to introduce a method that utilizes partial isotope profiles, only two mass isotopomers, to compute protein turnover rate. The use of partial isotope profiles alleviates the interferences from co-eluting contaminants in complex proteome mixtures and improves the accuracy of the estimation of label enrichment. In five different datasets, the technique consistently doubles the number of peptides with high goodness-of-fit characteristics of the turnover rate model. We also introduce a software tool, d2ome+, which automates the protein turnover estimation from partial isotope profiles.

Retention Time Alignment for Protein Turnover Studies Using Heavy Water Metabolic Labeling.

Retention time (RT) alignment has been important for robust protein identification and quantification in proteomics. In data-dependent acquisition mode, whereby the precursor ions are semistochastically chosen for fragmentation in MS/MS, the alignment is used in an approach termed matched between runs (MBR). MBR transfers peptides, which were fragmented and identified in one experiment, to a replicate experiment where they were not identified. Before the MBR transfer, the RTs of experiments are aligned to reduce the chance of erroneous transfers. Despite its widespread use in other areas of quantitative proteomics, RT alignment has not been applied in data analyses for protein turnover using an atom-based stable isotope-labeling agent such as metabolic labeling with deuterium oxide, D2O. Deuterium incorporation changes isotope profiles of intact peptides in full scans and their fragment ions in tandem mass spectra. It reduces the peptide identification rates in current database search engines. Therefore, the MBR becomes more important. Here, we report on an approach to incorporate RT alignment with peptide quantification in studies of proteome turnover using heavy water metabolic labeling and LC-MS. The RT alignment uses correlation-optimized time warping. The alignment, followed by the MBR, improves labeling time point coverage, especially for long labeling durations.

Software Tool for Visualization and Validation of Protein Turnover Rates Using Heavy Water Metabolic Labeling and LC-MS.

Metabolic stable isotope labeling followed by liquid chromatography coupled with mass spectrometry (LC-MS) is a powerful tool for in vivo protein turnover studies of individual proteins on a large scale and with high throughput. Turnover rates of thousands of proteins from dozens of time course experiments are determined by data processing tools, which are essential components of the workflows for automated extraction of turnover rates. The development of sophisticated algorithms for estimating protein turnover has been emphasized. However, the visualization and annotation of the time series data are no less important. The visualization tools help to validate the quality of the model fits, their goodness-of-fit characteristics, mass spectral features of peptides, and consistency of peptide identifications, among others. Here, we describe a graphical user interface (GUI) to visualize the results from the protein turnover analysis tool, d2ome, which determines protein turnover rates from metabolic D2O labeling followed by LC-MS. We emphasize the specific features of the time series data and their visualization in the GUI. The time series data visualized by the GUI can be saved in JPEG format for storage and further dissemination.

Prediction of Type 2 Diabetes Based on Machine Learning Algorithm.

Prediction of type 2 diabetes (T2D) occurrence allows a person at risk to take actions that can prevent onset or delay the progression of the disease. In this study, we developed a machine learning (ML) model to predict T2D occurrence in the following year (Y + 1) using variables in the current year (Y). The dataset for this study was collected at a private medical institute as electronic health records from 2013 to 2018. To construct the prediction model, key features were first selected using ANOVA tests, chi-squared tests, and recursive feature elimination methods. The resultant features were fasting plasma glucose (FPG), HbA1c, triglycerides, BMI, gamma-GTP, age, uric acid, sex, smoking, drinking, physical activity, and family history. We then employed logistic regression, random forest, support vector machine, XGBoost, and ensemble machine learning algorithms based on these variables to predict the outcome as normal (non-diabetic), prediabetes, or diabetes. Based on the experimental results, the performance of the prediction model proved to be reasonably good at forecasting the occurrence of T2D in the Korean population. The model can provide clinicians and patients with valuable predictive information on the likelihood of developing T2D. The cross-validation (CV) results showed that the ensemble models had a superior performance to that of the single models. The CV performance of the prediction models was improved by incorporating more medical history from the dataset.