Seqanswers Leaderboard Ad

Collapse

Announcement

Collapse
No announcement yet.
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Different DESeq2 results (both FC and pvalue) by changing col/row orders in input

    I'm getting different p-values and FCs from DESeq2 (v1.34.0) by simply changing the order of the columns in the count table. I can't understand why. I made sure the cols of count table match the rows of colData in both.

    The differences in this dataset are small ([-0.003, 0.003] for p-value, and [-0.002, 0.001] for FC). But in some other datasets, the difference of log2FC can be up to 0.8 (and the baseMean for that gene is 300), which is concerning.

    I tried setting `betaPrior = FALSE` but still got the difference.

    Column `Genotype` in colData is the only tested variable in the model. It contains `WT` and `Mutant` as factors, with `WT` as the 1st level.

    Actually if I change column `Genotype` to character type and rerun DESeq2 (so DESeq2 will convert them to factors with the warning message), I get the same results independent of the col/row orders in the input. So it seems the difference is somehow related to the factorization of the columns in colData, but I don't know why, and which results I should trust.



    Code:
    data <- readRDS('data.rds')
    
    dds1 <- DESeqDataSetFromMatrix(countData = data[['counts']], colData = data[['colData']], design = ~ Genotype)
    dds1 <- dds1[rowSums(counts(dds1)) > 1, ]
    dds1 <- DESeq(dds1, betaPrior = TRUE)
    res1 <- as.data.frame(results(dds1, contrast = list("GenotypeMutant","GenotypeWT")))
    
    # only reorder the inputs, but the columns of counts still match the rows of colData:
    dds2 <- DESeqDataSetFromMatrix(countData = data[['counts']][,c(4:6,1:3)], colData = data[['colData']][c(4:6,1:3),], design = ~ Genotype)
    dds2 <- dds2[rowSums(counts(dds2)) > 1, ]
    dds2 <- DESeq(dds2, betaPrior = TRUE)
    res2 <- as.data.frame(results(dds2, contrast = list("GenotypeMutant","GenotypeWT")))
    
    res1['ENSMUSG00000069515',]
    res2['ENSMUSG00000069515',]


    Code:
    baseMean log2FoldChange lfcSE stat pvalue
    ENSMUSG00000069515 3043.921 1.142537 0.09271227 12.32347 6.771949e-35
    padj
    ENSMUSG00000069515 4.605602e-31
    
    
    baseMean log2FoldChange lfcSE stat pvalue
    ENSMUSG00000069515 3043.921 1.142512 0.09272373 12.32168 6.923947e-35
    padj
    ENSMUSG00000069515 4.708976e-31

Latest Articles

Collapse

  • seqadmin
    Current Approaches to Protein Sequencing
    by seqadmin


    Proteins are often described as the workhorses of the cell, and identifying their sequences is key to understanding their role in biological processes and disease. Currently, the most common technique used to determine protein sequences is mass spectrometry. While still a valuable tool, mass spectrometry faces several limitations and requires a highly experienced scientist familiar with the equipment to operate it. Additionally, other proteomic methods, like affinity assays, are constrained...
    04-04-2024, 04:25 PM
  • seqadmin
    Strategies for Sequencing Challenging Samples
    by seqadmin


    Despite advancements in sequencing platforms and related sample preparation technologies, certain sample types continue to present significant challenges that can compromise sequencing results. Pedro Echave, Senior Manager of the Global Business Segment at Revvity, explained that the success of a sequencing experiment ultimately depends on the amount and integrity of the nucleic acid template (RNA or DNA) obtained from a sample. “The better the quality of the nucleic acid isolated...
    03-22-2024, 06:39 AM

ad_right_rmr

Collapse

News

Collapse

Topics Statistics Last Post
Started by seqadmin, 04-11-2024, 12:08 PM
0 responses
25 views
0 likes
Last Post seqadmin  
Started by seqadmin, 04-10-2024, 10:19 PM
0 responses
28 views
0 likes
Last Post seqadmin  
Started by seqadmin, 04-10-2024, 09:21 AM
0 responses
24 views
0 likes
Last Post seqadmin  
Started by seqadmin, 04-04-2024, 09:00 AM
0 responses
52 views
0 likes
Last Post seqadmin  
Working...
X