Cardiovascular Signature Temporal Clustering Platform

During cardiovascular disease progression, multiple systems in the myocardium (e.g., cardiac proteome) undergo diverse molecular changes. The temporal patterns of individual molecules depict their unique responses towards pathological drivers and contribute to underlying pathogenesis. Advances in high-throughput omics technology have enabled cost-effective temporal profiling of targeted systems in animal models of human disease. However, unsupervised analysis of temporal patterns from omics data remains challenging. Particularly, development of bioinformatic pipelines and unsupervised statistical methods are needed to accelerate mining biomedical insights from these complex datasets.

CV.Signature.TCP is a open source R package to identifying temporal molecular signatures from time-course omics data. It is consisted of 3 modules, Preprocessing, Clustering, and Evaluation. First, we preprocessed the temporal data using cubic splines or Principal Component Analysis (PCA), which accomplish both missing data imputation and denoising. Second, we created an unsupervised classification by K-means or hierarchical clustering. Third, we evaluate and identify biological entities (e.g., proteins) that are strongly related to temporal patterns.

We found that our platform produced biological meaningful clusters, enabling further functional delineation. Its flexible parameter settings and analytical routes allow a broader adaptation to other time-course omics data.

Installation

To install this package from GitHub, please use devtools and also set repositories to both CRAN and Bioconductor:

install.packages("devtools")
library("devtools")

setRepositories(ind=1:2) # select both CRAN and Bioconductor
install_github("UCLA-BD2K/CV.Signature.TCP")

Bioconductor

Some of Bioconductor dependencies may fail to be automatically installed. Therefore, if you get an error message about "package or namespace load failed", install them manually:

source("https://bioconductor.org/biocLite.R")
biocLite(c('multtest'))

XQuartz

On MacOS, you may have to install XQuartz (https://xquartz.org) to load rgl.

Reference

Forthcoming publication

Data

J Wang, H Choi, NC Chung, Q Cao, DCM Ng, B Mirza, SB Scruggs, D Wang, AO Garlid, P Ping (2018). Integrated dissection of the cysteine oxidative post-translational modification proteome during cardiac hypertrophy. Journal of Proteome Research https://pubs.acs.org/doi/abs/10.1021/acs.jproteome.8b00372