Error using DESeq2's nbinomLRT
1
0
Entering edit mode
guyho ▴ 20
@guyho-15677
Last seen 4 months ago
Israel

Hi,

I have a 2 factor experiment (2 mutations:A,B), with 2 levels (+/-) in each factor. All together there 4 conditions: wt, A, B, and A+B , with two replicates for each condition. I tried to use LRT test to isolate genes that are correspond to the mutations without interaction. the full model I used is ~A+B+A:B and the reduced is ~A:B I received the following error:

Error in nbinomLRT(object, full = full, reduced = reduced, quiet = quiet,  : 
  less than one degree of freedom, perhaps full and reduced models are not in the correct order

why are the degrees of freedom less than 1? Below I pasted the sessionInfo output. Thanks in advance.

Guy

R version 4.0.2 (2020-06-22)
Platform: x86_64-w64-mingw32/x64 (64-bit)
Running under: Windows 10 x64 (build 18362)

Matrix products: default

Random number generation:
 RNG:     Mersenne-Twister 
 Normal:  Inversion 
 Sample:  Rounding 

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

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

other attached packages:
 [1] ggplotify_0.0.5             gridExtra_2.3              
 [3] dplyr_1.0.0                 DESeq2_1.28.1              
 [5] SummarizedExperiment_1.18.1 DelayedArray_0.14.0        
 [7] matrixStats_0.56.0          Biobase_2.48.0             
 [9] GenomicRanges_1.40.0        GenomeInfoDb_1.24.2        
[11] IRanges_2.22.2              S4Vectors_0.26.1           
[13] BiocGenerics_0.34.0         ggplot2_3.3.2              
[15] readxl_1.3.1                data.table_1.12.8          

loaded via a namespace (and not attached):
 [1] Rcpp_1.0.4.6           locfit_1.5-9.4        
 [3] lattice_0.20-41        assertthat_0.2.1      
 [5] digest_0.6.25          R6_2.4.1              
 [7] cellranger_1.1.0       RSQLite_2.2.0         
 [9] pillar_1.4.4           zlibbioc_1.34.0       
[11] rlang_0.4.6            rstudioapi_0.11       
[13] annotate_1.66.0        blob_1.2.1            
[15] Matrix_1.2-18          labeling_0.3          
[17] splines_4.0.2          BiocParallel_1.22.0   
[19] geneplotter_1.66.0     pheatmap_1.0.12       
[21] RCurl_1.98-1.2         bit_1.1-15.2          
[23] munsell_0.5.0          compiler_4.0.2        
[25] pkgconfig_2.0.3        gridGraphics_0.5-0    
[27] tidyselect_1.1.0       tibble_3.0.1          
[29] GenomeInfoDbData_1.2.3 XML_3.99-0.4          
[31] fansi_0.4.1            crayon_1.3.4          
[33] withr_2.2.0            bitops_1.0-6          
[35] xtable_1.8-4           gtable_0.3.0          
[37] lifecycle_0.2.0        DBI_1.1.0             
[39] magrittr_1.5           scales_1.1.1          
[41] cli_2.0.2              farver_2.0.3          
[43] XVector_0.28.0         genefilter_1.70.0     
[45] ellipsis_0.3.1         rvcheck_0.1.8         
[47] generics_0.0.2         vctrs_0.3.1           
[49] RColorBrewer_1.1-2     tools_4.0.2           
[51] bit64_0.9-7            glue_1.4.1            
[53] purrr_0.3.4            survival_3.1-12       
[55] AnnotationDbi_1.50.1   colorspace_1.4-1      
[57] BiocManager_1.30.10    memoise_1.1.0
deseq2 LRT • 171 views
ADD COMMENT
0
Entering edit mode
@mikelove
Last seen 15 hours ago
United States

The reduced model is the null model, but here you've removed the main effects and left only the interaction. Take a look at the LRT section of the vignette.

ADD COMMENT
0
Entering edit mode

Thanks for the help.

So to find genes affected only by mutation A (and not by mutation B or interaction) should I: 1)test full=~A+B+A:B, reduced=~1 2)test full=~A+B+A:B, reduced=~A+B 3) test full=~A+B, reduced=~B

Select genes significant in 1 and not significant in 2 and 3?

ADD REPLY
0
Entering edit mode

I'd recommend collaborating with a statistician to work out the right statistical design for your experiment. We have a lot of material in the vignette, but beyond that I unfortunately don't have time to provide statistical consultation on the support site.

ADD REPLY

Login before adding your answer.

Traffic: 199 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