DESeq2 extracting relevant contrasts with interactions
0
Entering edit mode
jd348 • 0
@user-24862
Last seen 3 days ago

Hello,

I am attempting to analyze expression differences between three genotypes of mice (1 wildtype, 2 mutated) at three different ages. I would like to compare expression at a particular age across the genotypes – for example, compare wildtype to d52 at age 6 months. I have attached my sample list for clarity:

sample list

However, because my model matrix is not full rank, I manually edited the matrix as per the vignette. I am still having trouble though because when I call resultsNames(dds), I am not getting the contrasts that I want.


ddsHTSeq <- DESeqDataSetFromHTSeqCount(sampleTable = sampletable,
                                       directory = directory,
                                       design = ~ 1)
ddsHTSeq_filtered <- ddsHTSeq[rowSums(counts(ddsHTSeq)) > 10, ]

custom.matrix <- model.matrix(~ Genotype + Age:Genotype, data = colData(ddsHTSeq_filtered))
all.zero <- apply(custom.matrix, 2, function(x) all(x==0))
idx <- which(all.zero)
custom.matrix <- custom.matrix[,-idx]

dds <- DESeq(ddsHTSeq_filtered, full = custom.matrix)
resultsNames(dds)

Output: "Intercept" "Genotypemdx" "GenotypeWT" "Genotyped52.Age6m" "Genotypemdx.Age6m" "GenotypeWT.Age6m" "Genotypemdx.Agep8" "GenotypeWT.Agep8"

I also attempted to extract contrasts using the following:

results(dds, contrast = c("Genotype", "WT", "mdx"))

But I receive the following error: only list- and numeric-type contrasts are supported for user-supplied model matrices

Any help would be greatly appreciated!!

sessionInfo( )
R version 3.6.2 (2019-12-12)
Platform: x86_64-apple-darwin15.6.0 (64-bit)
Running under: macOS Mojave 10.14

Matrix products: default
BLAS:   /System/Library/Frameworks/Accelerate.framework/Versions/A/Frameworks/vecLib.framework/Versions/A/libBLAS.dylib
LAPACK: /Library/Frameworks/R.framework/Versions/3.6/Resources/lib/libRlapack.dylib

locale:
[1] en_US.UTF-8/en_US.UTF-8/en_US.UTF-8/C/en_US.UTF-8/en_US.UTF-8

attached base packages:
[1] parallel  stats4    stats     graphics  grDevices utils     datasets  methods   base     

other attached packages:
 [1] forcats_0.5.1               stringr_1.4.0               dplyr_1.0.4                 purrr_0.3.4                
 [5] readr_1.4.0                 tidyr_1.1.2                 tibble_3.0.6                tidyverse_1.3.0            
 [9] ggplot2_3.3.3               pheatmap_1.0.12             apeglm_1.8.0                DESeq2_1.26.0              
[13] SummarizedExperiment_1.16.1 DelayedArray_0.12.3         BiocParallel_1.20.1         matrixStats_0.58.0         
[17] Biobase_2.46.0              GenomicRanges_1.38.0        GenomeInfoDb_1.22.1         IRanges_2.20.2             
[21] S4Vectors_0.24.4            BiocGenerics_0.32.0        

loaded via a namespace (and not attached):
 [1] colorspace_2.0-0       ellipsis_0.3.1         htmlTable_2.1.0        XVector_0.26.0         base64enc_0.1-3       
 [6] fs_1.5.0               rstudioapi_0.13        farver_2.0.3           bit64_4.0.5            AnnotationDbi_1.48.0  
[11] mvtnorm_1.1-1          lubridate_1.7.9.2      xml2_1.3.2             splines_3.6.2          cachem_1.0.4          
[16] geneplotter_1.64.0     knitr_1.31             Formula_1.2-4          jsonlite_1.7.2         broom_0.7.4           
[21] annotate_1.64.0        cluster_2.1.1          dbplyr_2.1.0           png_0.1-7              compiler_3.6.2        
[26] httr_1.4.2             backports_1.2.1        assertthat_0.2.1       Matrix_1.3-2           fastmap_1.1.0         
[31] cli_2.3.0              htmltools_0.5.1.1      tools_3.6.2            coda_0.19-4            gtable_0.3.0          
[36] glue_1.4.2             GenomeInfoDbData_1.2.2 Rcpp_1.0.6             bbmle_1.0.23.1         cellranger_1.1.0      
[41] vctrs_0.3.6            xfun_0.21              rvest_0.3.6            lifecycle_1.0.0        XML_3.99-0.3          
[46] zlibbioc_1.32.0        MASS_7.3-53.1          scales_1.1.1           hms_1.0.0              RColorBrewer_1.1-2    
[51] memoise_2.0.0          gridExtra_2.3          emdbook_1.3.12         bdsmatrix_1.3-4        rpart_4.1-15          
[56] latticeExtra_0.6-29    stringi_1.5.3          RSQLite_2.2.3          genefilter_1.68.0      checkmate_2.0.0       
[61] rlang_0.4.10           pkgconfig_2.0.3        bitops_1.0-6           lattice_0.20-41        labeling_0.4.2        
[66] htmlwidgets_1.5.3      bit_4.0.4              tidyselect_1.1.0       plyr_1.8.6             magrittr_2.0.1        
[71] R6_2.5.0               generics_0.1.0         Hmisc_4.4-2            DBI_1.1.1              pillar_1.4.7          
[76] haven_2.3.1            foreign_0.8-75         withr_2.4.1            survival_3.2-7         RCurl_1.98-1.2        
[81] nnet_7.3-15            modelr_0.1.8           crayon_1.4.1           jpeg_0.1-8.1           locfit_1.5-9.4        
[86] grid_3.6.2             readxl_1.3.1           data.table_1.13.6      blob_1.2.1             reprex_1.0.0          
[91] digest_0.6.27          xtable_1.8-4           numDeriv_2016.8-1.1    munsell_0.5.0
DESeq2 DifferentialExpression • 90 views
ADD COMMENTlink
1
Entering edit mode
@mikelove
Last seen 2 minutes ago
United States

I'd recommend you work with a statistician to figure out how to map from contrasts of interest to the coefficients of the model matrix.

"only list- and numeric-type contrasts" this refers to how to provide information to contrast, and these are explained in ?results.

A list-type contrast looks like list("...", "...") where you can fill in the coefficient name for the numerator and for the denominator of the fold change, respectively.

If you are interested in a single coefficient, you should use name when calling results().

ADD COMMENTlink
0
Entering edit mode

Thanks for the response! Would using the following results(dds, contrast = list(c("GenotypeWT.Age6m","Genotypemdx.Age6m"))) contrast the differences in the WT genotype and mdx genotype for the 6 month old cohort?

ADD REPLYlink
1
Entering edit mode

I'm sorry I just don't have sufficient time to help with interpretation of results on the support site, and have to restrict myself to software related questions. I'd recommend to work with statistician to interpret and build the contrasts of interest.

ADD REPLYlink

Login before adding your answer.

Similar Posts
Loading Similar Posts
Traffic: 217 users visited in the last hour
Help About
FAQ
Access RSS
API
Stats

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.

Powered by the version 2.3.4