DESeq2 relevel question and resultsNames
1
0
Entering edit mode
Hannah • 0
@anna-11867
Last seen 2.6 years ago
United States

I am having trouble understanding why my resultsNames are missing comparisons that I thought should be there, as well as if I should relevel for both factors in my analysis?

I have two treatments 1-Year of sample and 2-Timpoint of collection. I am conducting an LRT test.

What I want to know is, shouldn't there be a "Year_H2019_vs_H1999", "Timepoint_T24_vs_T48" , "YearH1999.TimepointT0", "YearH2019.TimepointT0", "YearH1995.TimepointT0", "YearH1995.TimepointT24", and a "YearH1995.TimepointT48", present in the results(Names)?

My second question is: I want to have comparisons based on timepoint 0, but do I also need to relevel on the "Year" factor?

Thanks for any guidance on these problems.

#Here is my metadata
metadata
      Year Timepoint    Group
gm10 H1999       T24 1999_T24
gm11 H1995       T24 1995_T24
gm12 H1995       T24 1995_T24
gm13 H1999       T24 1999_T24
gm14 H1999       T48 1999_T48
gm15 H2019       T48 2019_T48
gm16 H1999       T48 1999_T48
gm17 H2019       T24 2019_T24
gm18 H1995       T24 1995_T24
gm19 H1999        T0  1999_T0
gm1  H2019       T24 2019_T24
gm20 H1995        T0  1995_T0
gm21 H1999        T0  1999_T0
gm22 H1999        T0  1999_T0
gm23 H1995        T0  1995_T0
gm24 H1999       T48 1999_T48
gm25 H2019       T48 2019_T48
gm26 H1999       T48 1999_T48
gm27 H1995       T48 1995_T48
gm28 H2019       T24 2019_T24
gm29 H1999        T0  1999_T0
gm2  H2019        T0  2019_T0
gm30 H1999       T24 1999_T24
gm31 H1995        T0  1995_T0
gm32 H2019        T0  2019_T0
gm33 H1999       T24 1999_T24
gm34 H2019       T48 2019_T48
gm35 H1995       T24 1995_T24
gm36 H2019       T24 2019_T24
gm3  H1995        T0  1995_T0
gm4  H1995       T48 1995_T48
gm5  H2019       T48 2019_T48
gm6  H1995       T48 1995_T48
gm7  H2019        T0  2019_T0
gm8  H1995       T48 1995_T48
gm9  H2019        T0  2019_T0

##My model 
dds<- DESeqDataSetFromMatrix(countData = mycounts, colData = metadata, design=~Year+Timepoint+Year:Timepoint)
##relevel for Timepoint 0 to be the reference
dds$Timepoint<- relevel(dds$Timepoint, ref= "T0")

dds<- DESeq(dds)
res <- results(dds)
> resultsNames(dds)
[1] "Intercept"              "Year_H1999_vs_H1995"    "Year_H2019_vs_H1995"    "Timepoint_T24_vs_T0"   
[5] "Timepoint_T48_vs_T0"    "YearH1999.TimepointT24" "YearH2019.TimepointT24" "YearH1999.TimepointT48"
[9] "YearH2019.TimepointT48"


# include your problematic code here with any corresponding output 
# please also include the results of running the following in an R session 

> sessionInfo( )
R version 4.0.1 (2020-06-06)
Platform: x86_64-w64-mingw32/x64 (64-bit)
Running under: Windows >= 8 x64 (build 9200)

Matrix products: default

locale:
[1] LC_COLLATE=English_United States.1252  LC_CTYPE=English_United States.1252   
[3] LC_MONETARY=English_United States.1252 LC_NUMERIC=C                          
[5] LC_TIME=English_United States.1252    

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               purrr_0.3.4                
 [4] readr_2.1.2                 tidyr_1.2.0                 tibble_3.1.6               
 [7] ggplot2_3.3.5               tidyverse_1.3.1             DESeq2_1.28.1              
[10] SummarizedExperiment_1.18.2 DelayedArray_0.14.1         matrixStats_0.61.0         
[13] Biobase_2.48.0              GenomicRanges_1.40.0        GenomeInfoDb_1.24.2        
[16] IRanges_2.22.2              S4Vectors_0.26.1            BiocGenerics_0.34.0        
[19] dplyr_1.0.8                

loaded via a namespace (and not attached):
 [1] httr_1.4.2             bit64_4.0.5            jsonlite_1.8.0         splines_4.0.1         
 [5] modelr_0.1.8           assertthat_0.2.1       blob_1.2.2             cellranger_1.1.0      
 [9] GenomeInfoDbData_1.2.3 yaml_2.2.2             pillar_1.7.0           RSQLite_2.2.10        
[13] backports_1.4.1        lattice_0.20-45        glue_1.6.1             RColorBrewer_1.1-3    
[17] XVector_0.28.0         rvest_1.0.2            colorspace_2.0-2       Matrix_1.4-0          
[21] XML_3.99-0.8           pkgconfig_2.0.3        broom_0.7.12           haven_2.4.3           
[25] genefilter_1.70.0      zlibbioc_1.34.0        xtable_1.8-4           scales_1.1.1          
[29] tzdb_0.2.0             BiocParallel_1.22.0    annotate_1.66.0        generics_0.1.2        
[33] ellipsis_0.3.2         withr_2.5.0            cachem_1.0.6           cli_3.1.1             
[37] survival_3.2-13        magrittr_2.0.2         crayon_1.5.1           readxl_1.3.1          
[41] memoise_2.0.1          fs_1.5.2               fansi_1.0.2            xml2_1.3.3            
[45] tools_4.0.1            hms_1.1.1              lifecycle_1.0.1        reprex_2.0.1          
[49] munsell_0.5.0          locfit_1.5-9.4         AnnotationDbi_1.50.3   compiler_4.0.1        
[53] rlang_1.0.1            grid_4.0.1             RCurl_1.98-1.6         rstudioapi_0.13       
[57] bitops_1.0-7           gtable_0.3.0           DBI_1.1.2              R6_2.5.1              
[61] lubridate_1.8.0        fastmap_1.1.0          bit_4.0.4              utf8_1.2.2            
[65] stringi_1.7.6          Rcpp_1.0.8             vctrs_0.3.8            geneplotter_1.66.0    
[69] dbplyr_2.1.1           tidyselect_1.1.2
DESeq2 • 956 views
ADD COMMENT
0
Entering edit mode
@mikelove
Last seen 10 hours ago
United States

I am having trouble understanding why my resultsNames are missing comparisons that I thought should be there, as well as if I should relevel for both factors in my analysis?

For questions about statistical design and interpretation of results, I recommend consulting a local statistician or someone familiar with linear models in R. I have to reserve my time on the support site for software related questions.

ADD COMMENT
0
Entering edit mode

I apologize, I thought this was a software question on releveling and missing comparisons. Maybe someone else could share a link that has a good example or documentation of a two factor analysis that includes releveling? I have been trying to find thorough documentation, but haven't been able to find a good example.

ADD REPLY
0
Entering edit mode

In R, one can specify a particular linear model using formula and this produces a set of coefficients: this is generic to all methods in Bioconductor and beyond. So anyone who knows how to construct and interpret the design and coefficients in R would be able to help you. I try to therefore reserve my time on the support site to specific questions about DESeq2 software. There are a lot of users online with general statistical consulting questions about how to analyze their datasets and interpret results, but I just don't have time to address all of those questions about experimental design here, while also maintaining and supporting software.

ADD REPLY

Login before adding your answer.

Traffic: 684 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 2.3.6