In nesscoder/TimeCycle: TimeCycle: Topology Inspired MEthod for the Detection of Cycling Transcripts

TimeCycle: Topology Inspired MEthod for the Detection of Cycling Transcripts in Circadian Time-Series Data

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>",
  fig.path = "man/figures/",
  out.width = "100%"
)
library(TimeCycle)
suppressMessages(library(tidyverse))
library(badger)

r badge_doi("10.1093/bioinformatics/btab476", "yellow") r badge_devel("nesscoder/TimeCycle", "blue") r badge_code_size("nesscoder/TimeCycle")

TimeCycle is designed to detect rhythmic genes in circadian transcriptomic time-series data. Based on topological data analysis, TimeCycle provides a reliable and efficent reference-free framework for cycle detection --- handling custom sampling schemes, replicates, and missing data.

• To learn more about the theory and usage of TimeCycle, see our video overview and following vignette("TimeCycle").
• For a comprehensive analysis and discussion of TimeCycle's performance in detecting rhythmic genes, see the accompanying paper.
• For details pertaining to the data and source code used in the analysis, see the nesscoder/TimeCycle-data git repository.

---

Installation

TimeCycle has not yet been published on Bioconductor. In the interim download the development version from GitHub.

# Install development version from GitHub
devtools::install_github("nesscoder/TimeCycle")

---

Usage

Circadian cycling detection can easily be achieved using TimeCycle's main function. Get started with TimeCycle() by defining the data and replicate labels.

In this example, we will use the zhang2014() gene expression set consists of mouse livers sampled every 2-h for 48-h with a single replicate (i.e. 24 time points). TimeCycle assumes a default circadian period of 24-h.

• See zhang2014() for additional information about the example data set.
• See replicate labels for additional information regarding repLabel.

suppressMessages(TimeCycle(data = zhang2014[1:3,], repLabel = rep(1,24), period = 24, resamplings = 10))

library(TimeCycle)

#set seed for reproducibility with random variables in example usage
set.seed(1234) 

TimeCycleResults <- TimeCycle(data = zhang2014, repLabel = rep(1,24))

Once TimeCycle has finished processing, simply check the output and filter for the genes of interest. In this example, we filter for genes with a period of oscillation between 22 and 26 hours and an FDR \< 0.05.

library(tidyverse)

TimeCycleResults %>%
  filter(22 < Period.in.Hours & Period.in.Hours < 26) %>%
  filter(pVals.adj < 0.05) %>%
  glimpse()

See vignette("TimeCycle") for a detailed description of algorithm design and suggestions for custom parameter selection.

nesscoder/TimeCycle documentation built on June 29, 2021, 5:16 a.m.