Gaussian Process Modeling of Protein Turnover

Mahbubur Rahman, Stephen F. Previs, Takhar Kasumov, Rovshan Sadygov

Research output: Contribution to journalArticle

8 Citations (Scopus)

Abstract

We describe a stochastic model to compute in vivo protein turnover rate constants from stable-isotope labeling and high-throughput liquid chromatography-mass spectrometry experiments. We show that the often-used one- and two-compartment nonstochastic models allow explicit solutions from the corresponding stochastic differential equations. The resulting stochastic process is a Gaussian processes with Ornstein-Uhlenbeck covariance matrix. We applied the stochastic model to a large-scale data set from 15N labeling and compared its performance metrics with those of the nonstochastic curve fitting. The comparison showed that for more than 99% of proteins, the stochastic model produced better fits to the experimental data (based on residual sum of squares). The model was used for extracting protein-decay rate constants from mouse brain (slow turnover) and liver (fast turnover) samples. We found that the most affected (compared to two-exponent curve fitting) results were those for liver proteins. The ratio of the median of degradation rate constants of liver proteins to those of brain proteins increased 4-fold in stochastic modeling compared to the two-exponent fitting. Stochastic modeling predicted stronger differences of protein turnover processes between mouse liver and brain than previously estimated. The model is independent of the labeling isotope. To show this, we also applied the model to protein turnover studied in induced heart failure in rats, in which metabolic labeling was achieved by administering heavy water. No changes in the model were necessary for adapting to heavy-water labeling. The approach has been implemented in a freely available R code.

Original languageEnglish (US)
Pages (from-to)2115-2122
Number of pages8
JournalJournal of Proteome Research
Volume15
Issue number7
DOIs
StatePublished - Jul 1 2016

Fingerprint

Labeling
Liver
Proteins
Stochastic models
Deuterium Oxide
Isotope Labeling
Rate constants
Brain
Curve fitting
Isotopes
Stochastic Processes
Liquid chromatography
Covariance matrix
Random processes
Liquid Chromatography
Mass spectrometry
Rats
Mass Spectrometry
Differential equations
Heart Failure

Keywords

  • dynamic proteome
  • Gaussian process
  • mass spectrometry
  • Ornstein-Uhlenbeck process
  • protein degradation rate constant
  • protein turnover rate constant
  • stable isotope labeling
  • stochastic differential equation for protein turnover rate constant

ASJC Scopus subject areas

  • Biochemistry
  • Chemistry(all)

Cite this

Gaussian Process Modeling of Protein Turnover. / Rahman, Mahbubur; Previs, Stephen F.; Kasumov, Takhar; Sadygov, Rovshan.

In: Journal of Proteome Research, Vol. 15, No. 7, 01.07.2016, p. 2115-2122.

Research output: Contribution to journalArticle

Rahman, Mahbubur ; Previs, Stephen F. ; Kasumov, Takhar ; Sadygov, Rovshan. / Gaussian Process Modeling of Protein Turnover. In: Journal of Proteome Research. 2016 ; Vol. 15, No. 7. pp. 2115-2122.
@article{d0aba835030f402fafc9dbafe3220ba5,
title = "Gaussian Process Modeling of Protein Turnover",
abstract = "We describe a stochastic model to compute in vivo protein turnover rate constants from stable-isotope labeling and high-throughput liquid chromatography-mass spectrometry experiments. We show that the often-used one- and two-compartment nonstochastic models allow explicit solutions from the corresponding stochastic differential equations. The resulting stochastic process is a Gaussian processes with Ornstein-Uhlenbeck covariance matrix. We applied the stochastic model to a large-scale data set from 15N labeling and compared its performance metrics with those of the nonstochastic curve fitting. The comparison showed that for more than 99{\%} of proteins, the stochastic model produced better fits to the experimental data (based on residual sum of squares). The model was used for extracting protein-decay rate constants from mouse brain (slow turnover) and liver (fast turnover) samples. We found that the most affected (compared to two-exponent curve fitting) results were those for liver proteins. The ratio of the median of degradation rate constants of liver proteins to those of brain proteins increased 4-fold in stochastic modeling compared to the two-exponent fitting. Stochastic modeling predicted stronger differences of protein turnover processes between mouse liver and brain than previously estimated. The model is independent of the labeling isotope. To show this, we also applied the model to protein turnover studied in induced heart failure in rats, in which metabolic labeling was achieved by administering heavy water. No changes in the model were necessary for adapting to heavy-water labeling. The approach has been implemented in a freely available R code.",
keywords = "dynamic proteome, Gaussian process, mass spectrometry, Ornstein-Uhlenbeck process, protein degradation rate constant, protein turnover rate constant, stable isotope labeling, stochastic differential equation for protein turnover rate constant",
author = "Mahbubur Rahman and Previs, {Stephen F.} and Takhar Kasumov and Rovshan Sadygov",
year = "2016",
month = "7",
day = "1",
doi = "10.1021/acs.jproteome.5b00990",
language = "English (US)",
volume = "15",
pages = "2115--2122",
journal = "Journal of Proteome Research",
issn = "1535-3893",
publisher = "American Chemical Society",
number = "7",

}

TY - JOUR

T1 - Gaussian Process Modeling of Protein Turnover

AU - Rahman, Mahbubur

AU - Previs, Stephen F.

AU - Kasumov, Takhar

AU - Sadygov, Rovshan

PY - 2016/7/1

Y1 - 2016/7/1

N2 - We describe a stochastic model to compute in vivo protein turnover rate constants from stable-isotope labeling and high-throughput liquid chromatography-mass spectrometry experiments. We show that the often-used one- and two-compartment nonstochastic models allow explicit solutions from the corresponding stochastic differential equations. The resulting stochastic process is a Gaussian processes with Ornstein-Uhlenbeck covariance matrix. We applied the stochastic model to a large-scale data set from 15N labeling and compared its performance metrics with those of the nonstochastic curve fitting. The comparison showed that for more than 99% of proteins, the stochastic model produced better fits to the experimental data (based on residual sum of squares). The model was used for extracting protein-decay rate constants from mouse brain (slow turnover) and liver (fast turnover) samples. We found that the most affected (compared to two-exponent curve fitting) results were those for liver proteins. The ratio of the median of degradation rate constants of liver proteins to those of brain proteins increased 4-fold in stochastic modeling compared to the two-exponent fitting. Stochastic modeling predicted stronger differences of protein turnover processes between mouse liver and brain than previously estimated. The model is independent of the labeling isotope. To show this, we also applied the model to protein turnover studied in induced heart failure in rats, in which metabolic labeling was achieved by administering heavy water. No changes in the model were necessary for adapting to heavy-water labeling. The approach has been implemented in a freely available R code.

AB - We describe a stochastic model to compute in vivo protein turnover rate constants from stable-isotope labeling and high-throughput liquid chromatography-mass spectrometry experiments. We show that the often-used one- and two-compartment nonstochastic models allow explicit solutions from the corresponding stochastic differential equations. The resulting stochastic process is a Gaussian processes with Ornstein-Uhlenbeck covariance matrix. We applied the stochastic model to a large-scale data set from 15N labeling and compared its performance metrics with those of the nonstochastic curve fitting. The comparison showed that for more than 99% of proteins, the stochastic model produced better fits to the experimental data (based on residual sum of squares). The model was used for extracting protein-decay rate constants from mouse brain (slow turnover) and liver (fast turnover) samples. We found that the most affected (compared to two-exponent curve fitting) results were those for liver proteins. The ratio of the median of degradation rate constants of liver proteins to those of brain proteins increased 4-fold in stochastic modeling compared to the two-exponent fitting. Stochastic modeling predicted stronger differences of protein turnover processes between mouse liver and brain than previously estimated. The model is independent of the labeling isotope. To show this, we also applied the model to protein turnover studied in induced heart failure in rats, in which metabolic labeling was achieved by administering heavy water. No changes in the model were necessary for adapting to heavy-water labeling. The approach has been implemented in a freely available R code.

KW - dynamic proteome

KW - Gaussian process

KW - mass spectrometry

KW - Ornstein-Uhlenbeck process

KW - protein degradation rate constant

KW - protein turnover rate constant

KW - stable isotope labeling

KW - stochastic differential equation for protein turnover rate constant

UR - http://www.scopus.com/inward/record.url?scp=84977083534&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84977083534&partnerID=8YFLogxK

U2 - 10.1021/acs.jproteome.5b00990

DO - 10.1021/acs.jproteome.5b00990

M3 - Article

VL - 15

SP - 2115

EP - 2122

JO - Journal of Proteome Research

JF - Journal of Proteome Research

SN - 1535-3893

IS - 7

ER -