Three-dimensional dose and LETD prediction in proton therapy using artificial neural networks.

PURPOSE: Challenges in proton therapy include identifying patients most likely to benefit; ensuring consistent, high-quality plans as its adoption becomes more widespread; and recognizing biological uncertainties that may be related to increased relative biologic effectiveness driven by linear energy transfer (LET). Knowledge-based planning (KBP) is a domain that may help to address all three. METHODS: Artificial neural networks were trained using 117 unique treatment plans and associated dose and dose-weighted LET (LETD ) distributions. The data set was split into training (n = 82), validation (n = 17), and test (n = 18) sets. Model performance was evaluated on the test set using dose- and LETD -volume metrics in the clinical target volume (CTV) and nearby organs at risk and Dice similarity coefficients (DSC) comparing predicted and planned isodose lines at 50%, 75%, and 95% of the prescription dose. RESULTS: Dose-volume metrics significantly differed (α = 0.05) between predicted and planned dose distributions in only one dose-volume metric, D2% to the CTV. The maximum observed root mean square (RMS) difference between corresponding metrics was 4.3 GyRBE (8% of prescription) for D1cc to optic chiasm. DSC were 0.90, 0.93, and 0.88 for the 50%, 75%, and 95% isodose lines, respectively. LETD -volume metrics significantly differed in all but one metric, L0.1cc of the brainstem. The maximum observed difference in RMS differences for LETD metrics was 1.0 keV/µm for L0.1cc to brainstem. CONCLUSIONS: We have devised the first three-dimensional dose and LETD -prediction model for cranial proton radiation therapy has been developed. Dose accuracy compared favorably with that of previously published models in other treatment sites. The agreement in LETD supports future investigations with biological doses in mind to enable the full potential of KBP in proton therapy.

JUMPt: Comprehensive Protein Turnover Modeling of In Vivo Pulse SILAC Data by Ordinary Differential Equations.

Recent advances in mass spectrometry (MS)-based proteomics allow the measurement of turnover rates of thousands of proteins using dynamic labeling methods, such as pulse stable isotope labeling by amino acids in cell culture (pSILAC). However, when applying the pSILAC strategy to multicellular animals (e.g., mice), the labeling process is significantly delayed by native amino acids recycled from protein degradation in vivo, raising a challenge of defining accurate protein turnover rates. Here, we report JUMPt, a software package using a novel ordinary differential equation (ODE)-based mathematical model to determine reliable rates of protein degradation. The uniqueness of JUMPt is to consider amino acid recycling and fit the kinetics of the labeling amino acid (e.g., Lys) and whole proteome simultaneously to derive half-lives of individual proteins. Multiple settings in the software are designed to enable simple to comprehensive data inputs for precise analysis of half-lives with flexibility. We examined the software by studying the turnover of thousands of proteins in the pSILAC brain and liver tissues. The results were largely consistent with the proteome turnover measurements from previous studies. The long-lived proteins are enriched in the integral membrane, myelin sheath, and mitochondrion in the brain. In summary, the ODE-based JUMPt software is an effective proteomics tool for analyzing large-scale protein turnover, and the software is publicly available on GitHub (https://github.com/JUMPSuite/JUMPt) to the research community.

Spectral Library Search Improves Assignment of TMT Labeled MS/MS Spectra.

Tandem mass tag (TMT)-based liquid chromatography-tandem mass spectrometry (LC-MS/MS) is a proven approach for large-scale multiplexed protein quantification. However, the identification of TMT-labeled peptides is compromised by the labeling during traditional sequence database searches. In this study, we aim to use a spectral library search to increase the sensitivity and specificity of peptide identification for TMT-based MS data. Compared to MS/MS spectra of unlabeled peptides, the spectra of TMT-labeled counterparts usually display intensified b ions, suggesting that TMT labeling can alter product ion patterns during MS/MS fragementation. We compiled a human TMT spectral library of 401,168 unique peptides of high quality from millions of peptide-spectrum matches in tens of profiling projects, matching to 14,048 nonredundant proteins (13,953 genes). A mouse TMT spectral library of similar size was also constructed. The libraries were subsequently appended with decoy spectra to evaluate the false discovery rate, which was validated by a simulated null TMT data set. The performance of the library search was further optimized by removing TMT reporter ions and selecting an appropriate library construction method. Finally, we searched a human TMT data set against the spectral library to demonstrate that the spectral library outperformed the sequence database. Both human and mouse TMT libraries were made publicly available to the research community.