Differentially expressed genes from Deseq2 have similar counts
Entering edit mode
daianeh • 0
Last seen 6 weeks ago
United States


I'm running an analysis with Deseq2 to find differentially expressed genes between two different disease conditions in the heart. It was hard to find what was separating the samples on the PCA, but it seems to be heart region on PCs 3 and 4:

enter image description here enter image description here

So I'm controlling for that on the design, together with age and ancestry:

design= ~ Sex + Age + Tissue_Location + Condition

I get as result:

out of 53147 with nonzero total read count adjusted p-value < 0.1 LFC > 0 (up) : 2290, 4.3% LFC < 0 (down) : 367, 0.69% outliers 1 : 0, 0% low counts 2 : 23699, 45%

But when I plot the counts of the gene with maximum fold change or smallest p-value, they look like they have very similar counts apart from a couple of outliers:

gene with largest fold change

gene with smallest pvalue

Does that mean there's something wrong with my analysis?

Also the gene with highest fold change has log2FoldChange = 5.8. Isn't that suspiciously high?

sessionInfo( )

R version 4.1.0 (2021-05-18)
Platform: x86_64-pc-linux-gnu (64-bit)
Running under: CentOS Linux 7 (Core)

Matrix products: default
BLAS/LAPACK: /hpc/packages/minerva-centos7/intel/parallel_studio_xe_2019/compilers_and_libraries_2019.0.117/linux/mkl/lib/intel64_lin/libmkl_gf_lp64.so

 [1] LC_CTYPE=en_US.UTF-8       LC_NUMERIC=C
 [3] LC_TIME=en_US.UTF-8        LC_COLLATE=en_US.UTF-8
 [7] LC_PAPER=en_US.UTF-8       LC_NAME=C
 [9] LC_ADDRESS=C               LC_TELEPHONE=C

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

other attached packages:
 [1] gplots_3.1.1                ggplot2_3.3.3
 [3] dplyr_1.0.6                 DESeq2_1.32.0
 [5] SummarizedExperiment_1.22.0 Biobase_2.52.0
 [7] MatrixGenerics_1.4.0        matrixStats_0.59.0
 [9] GenomicRanges_1.44.0        GenomeInfoDb_1.28.0
[11] IRanges_2.26.0              S4Vectors_0.30.0
[13] BiocGenerics_0.38.0

loaded via a namespace (and not attached):
 [1] locfit_1.5-9.4         Rcpp_1.0.8.3           lattice_0.20-44
 [4] gtools_3.8.2           png_0.1-7              Biostrings_2.60.1
 [7] assertthat_0.2.1       utf8_1.2.1             R6_2.5.0
[10] RSQLite_2.2.7          httr_1.4.2             pillar_1.6.1
[13] zlibbioc_1.38.0        rlang_0.4.11           rstudioapi_0.13
[16] annotate_1.70.0        blob_1.2.1             Matrix_1.3-4
[19] splines_4.1.0          BiocParallel_1.26.0    geneplotter_1.70.0
[22] RCurl_1.98-1.3         bit_4.0.4              munsell_0.5.0
[25] DelayedArray_0.18.0    compiler_4.1.0         pkgconfig_2.0.3
[28] tidyselect_1.1.1       KEGGREST_1.32.0        tibble_3.1.2
[31] GenomeInfoDbData_1.2.6 XML_3.99-0.6           fansi_0.5.0
[34] withr_2.4.2            crayon_1.4.1           bitops_1.0-7
[37] grid_4.1.0             xtable_1.8-4           gtable_0.3.0
[40] lifecycle_1.0.0        DBI_1.1.1              magrittr_2.0.3
[43] datasets_4.1.0         scales_1.1.1           KernSmooth_2.23-20
[46] cachem_1.0.5           XVector_0.32.0         genefilter_1.74.0
[49] ellipsis_0.3.2         vctrs_0.3.8            generics_0.1.0
[52] RColorBrewer_1.1-2     tools_4.1.0            bit64_4.0.5
[55] glue_1.4.2             purrr_0.3.4            fastmap_1.1.0
[58] survival_3.2-11        AnnotationDbi_1.54.0   colorspace_2.0-1
[61] caTools_1.18.2         memoise_2.0.0
DESeq2 • 105 views
Entering edit mode
Basti ▴ 440
Last seen 2 hours ago

You will get your answer here : Contradictory results in smallRNA differential expression analysis using DESeq2


Login before adding your answer.

Traffic: 474 users visited in the last hour
Help About
Access RSS

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

Powered by the version 2.3.6