A Snakemake profile for running workflows on the University of Cambridge CSD3 HPC system using the SLURM executor.
The profile is intended to allow Snakemake to be run from a CSD3 login node while submitting production jobs to the SLURM queue.
Copy config.yaml into your Snakemake profile directory, then edit the CSD3 resource settings:
| #' Create numeric contrast for DESeq2::results() function | |
| #' | |
| #' @param object a DESeqDataSet object. | |
| #' @param group1 logical expression indicating the columns from colData(object) that form part of group1 (numerator). | |
| #' @param group2 logical expression indicating the columns from colData(object) that form part of group2 (denominator). | |
| #' | |
| #' @examples | |
| #' # example data for a two-factor design with interaction | |
| #' dds <- makeExampleDESeqDataSet() | |
| #' dds$timepoint <- factor(rep(1:3, 4)) |
| #' Calculate relative change across time | |
| #' | |
| #' @param value a vector with the values to calculate the change over. | |
| #' @param time a vector with the time points. | |
| #' @param ref_time the reference time point to scale the change by. | |
| relChange <- function(value, time, ref_time = min(time)){ | |
| # check if reference time is contained in the given vector | |
| if(any(is.na(value))) stop("Missing values in 'value' not supported. Remove them first.") | |
| if(any(is.na(time))) stop("Missing values in 'time' not supported. Remove them first.") |
Helper functions to manipulate and/or coerce qtl2 objects. Eventually might make sense to submit as pull request?
| #' Tidying methods for DESeq2 DESeqTransform objects | |
| #' | |
| #' This reshapes a DESeq2 transform object (output from `rlog()` and | |
| #' `varianceStabilizingTransform()` functions) into a tidy format. | |
| #' | |
| #' @param x DESeqTransform object | |
| #' @param colData whether colData should be included in the tidied output | |
| #' for those in the DESeqTransform object. | |
| #' @param ... extra arguments (not used) | |
| #' |
| #' Perform Principal Components Analysis on a DESeqTransform object | |
| #' | |
| #' This function is based on the `DESeq2::plotPCA()` function, but returns the | |
| #' results of `prcomp` in a tidy list format. This is more flexible for further | |
| #' custom plotting and exploring factor loadings of the PCA. | |
| #' | |
| #' @param x an object of class DESeqTransform | |
| #' @ntop number of most-variable genes to select. Igored if "genes" is specified. | |
| #' @genes character vector of specific genes to use | |
| #' |