README.md

SwipProteomics

This repository is an archive of the proteomics data and source code used in the analyses by Courtland et al., 2021.

Please see soderling-lab/SwipProteomics for an extant version of this code.

wash-module

Usage Example

# download the repository as an R package
devtools::install_github("twesleyb/SwipProteomics")

library(dplyr)
library(SwipProteomics)

# load the normalized data
data(swip_tmt)

# washc4's uniprot ID
data(swip)

# fit a model
fx <- log2(Intensity) ~ 0 + Condition + (1|Mixture)
fm <- lmerTest::lmer(fx,data=swip_tmt %>% subset(Protein==swip))

# create a contrast
LT <- getContrast(fm,"Mutant","Control")

# assess contrast 
res <- lmerTestContrast(fm, LT) %>% mutate(Contrast='Mutant-Control') %>% unique()

knitr::kable(res)

|Contrast | log2FC| percentControl| SE| Tstatistic| Pvalue| DF| S2|isSingular | |:--------------|---------:|--------------:|---------:|----------:|------:|--:|--------:|:----------| |Mutant-Control | -1.401866| 0.3784393| 0.0264791| -52.94235| 0| 28| 0.007362|TRUE |

lmerTestContrast returns a data.frame with statistics from the model-based contrast. The column isSingular=TRUE in this case indicates that the variance attributes to Mixture is negligible.


## fit WASH Complex

library(dplyr)
library(SwipProteomics)

data(washc_prots)

# module-level model includes ranef Protein
fx1 <- log2(rel_Intensity) ~ 0 + Condition + (1|Protein)

# fit the model
fm1 <- lmerTest::lmer(fx1, data = swip_tmt %>% subset(Protein %in% washc_prots))

# assess overall 'Mutant-Control' comparison
res1 <- lmerTestContrast(fm1, LT) %>% mutate(Contrast='Mutant-Control') %>% unique()

knitr::kable(res)

|Contrast | log2FC| percentControl| SE| Tstatistic| Pvalue| DF| S2|isSingular | |:--------------|---------:|--------------:|---------:|----------:|------:|---:|---------:|:----------| |Mutant-Control | -1.379633| 0.3843165| 0.0392109| -35.18497| 0| 151| 0.0645747|FALSE |



twesleyb/SwipProteomics documentation built on Sept. 15, 2021, 7:38 a.m.