No Evidence for Sex-Differential Transcriptomes Driving Genome-Wide Sex-Differential Natural Selection
Matthew J. Ming, Changde Cheng, Mark Kirkpatrick, Arbel Harpak
Here are the code and scripts used to produce the analysis for "No Evidence for Sex-Differential Transcriptomes Driving Genome-Wide Sex-Differential Natural Selection"
- All code can be found in the main directory. There are no subdirectories
- The following software were used:
- The following R packages were used:
- library(ggplot2)
- library(ggpubr)
- library(vroom)
- library(ggrepel)
- library(grid)
- library(pBrackets)
Main code is contained in R scripts. The Slurm script listed as script 2 is a wrapper which run R code over multiple files (e.g., multiple ancestries or tissue types). The slurm file is listed with a number (2) and the R scripts run by this slurm file is listed as number and letter (2a). Scripts 3 and 4 each have 2 versions, one for all non-gonad tissues and one for gonads. The gonad version is listed as 3a and 4a respectively.
All final figures were produced using TwinPeaksFigs_Rev_FINAL.R
.
Get original Twin Peaks curve from 1000 Genomes data (Fig 2)
Get Twin Peaks curves from sex label randomization in GTEx over multiple iterations (Fig 2)
R script called by ChengRepRand_GTExONLY.slurm
for getting Twin Peaks sex label randomization (Fig 2)
Get A-values for updated linear regression method analysis in all tissues (except gonads) (Fig S1)
Same as NewAnalysis_eQTLRegression_MH.R
, but modified slightly for the gonads specifically (Fig S1)
Get A-values using an improved weighted Standard Major Axis regression method (Fig 3)
Same as NewAnalysis_eQTLRegression_MH_Revision.R
, but modified slightly for the gonads specifically (Fig 3)
Final script for composing both main text and supplemental figures (Figs 1-3, S1). Depends on output from all previous scripts