Microarray DGE analysis

Question

[bioinformatics]

Hi,

https://www.ncbi.nlm.nih.gov/geo/geo2r/?acc=GSE30929

I have performed DGE analysis between WDLPS and DDLPS tumour samples in R and have the following results table. I ended up with the same results as the GEO2R.

         logFC AveExpr   t P.Value adj.P.Val  B
207175_at    -5.7     8.8 -14 3.1e-25   7.0e-21 47
213706_at    -3.5     8.9 -14 1.0e-24   1.1e-20 46
203548_s_at  -4.7    10.1 -14 2.8e-24   2.1e-20 45
203549_s_at  -4.3    10.1 -14 9.0e-24   4.4e-20 43
205913_at    -4.5     9.9 -14 9.9e-24   4.4e-20 43
218736_s_at  -2.5     8.7 -13 4.3e-23   1.6e-19 42
203571_s_at  -4.0     9.3 -13 2.0e-22   6.3e-19 40
49452_at     -3.0     9.1 -13 4.3e-22   1.2e-18 40
43427_at     -2.8     8.9 -13 5.3e-22   1.3e-18 39
219398_at    -3.3     9.6 -13 9.3e-22   2.1e-18 39

However, does the logFC of -5.7 in the first row of the table indicate that the gene is unregulated or down regulated in WDLPS or DDLPS?

Thanks!

Answer 1

[bk11]

It will depend upon how did you defined two groups. Would you please tell which is your control and which is test?

[bioinformatics]

I am uncertain. Below is my r code:

> library(GEOquery)
> my_id <- "GSE30929"
> gse <- getGEO(my_id)
> length(gse)
[1] 1
> gse <-gse[[1]]
> gse
gs0 <- pData(gse)$`subtype:ch1`   
> table(gs0)
gs0
 dedifferentiated              myxoid   myxoid/round cell 
                 40                  17                  11 
        pleomorphic well-differentiated 
                 20                  52 
> 
> keep <- c("dedifferentiated", "well-differentiated")
> sample_idx <- which(gs0 %in% keep)
> 
> gse <- gse[, sample_idx]
> 
> gs <- pData(gse)$`subtype:ch1`
> table(gs)
gs
   dedifferentiated well-differentiated 
                 40                  52

> pData(gse)$data_processing[1]
summary(exprs(gse))
 library(dplyr)
> sampleInfo <- pData(gse)
> head(sampleInfo)

table(sampleInfo$characteristics_ch1.1)

 sampleInfo <- select(sampleInfo, characteristics_ch1.1)
> sampleInfo <- rename(sampleInfo, sample = characteristics_ch1.1)
> 
> head(sampleInfo)
                     dim(sampleInfo)
[1] 92  1
> sampleInfo$sample
sampleInfo <- sampleInfo[,-1]
> sampleInfo
library(limma)
> design <- model.matrix(~0 + sampleInfo$group)
design
cutoff <- median(exprs(gse))
> 
> ## TRUE or FALSE for whether each gene is "expressed" in each sample
> is_expressed <- exprs(gse) > cutoff
> 
> ## Identify genes expressed in more than 2 samples
> 
> keep <- rowSums(is_expressed) > 3
> 
> ## check how many genes are removed / retained.
> table(keep)
 ## subset to just those expressed genes
> gse <- gse[keep,]
> aw <- arrayWeights(exprs(gse),design)
> aw
fit <- lmFit(exprs(gse), design,
+              weights = aw)
head(fit$coefficients)
fit2 <- contrasts.fit(fit, contrasts)
fit2 <- eBayes(fit2)
topTable(fit2)
top_genes <- topTable(fit2, coef = 1, n = 2000)
write.csv(top2000, "dge_data.csv", row.names = FALSE)

[bioinformatics]

I am not sure, I think DDLPS might be the test and WDLPS the control. Is this correct? I just performed a DGE analyses between the two tumour subtypes.

[bk11]

Please show me your design table. It looks like you are comparing between subtype: dedifferentiated vs subtype: well-differentiated.

[bioinformatics]

    head(design)
      sample id: ALT476 sample id: ALT479 sample id: ALT486 sample id: ALT497
    1                 0                 0                 0                 0
    2                 0                 0                 0                 0
    3                 0                 0                 0                 0
    4                 0                 0                 0                 0
    5                 0                 0                 0                 0
    6                 0                 0                 0                 0
      sample id: ALT500 sample id: DD0728 sample id: DD103 sample id: DD112
    1                 0                 1                0                0
    2                 0                 0                1                0
    3                 0                 0                0                1
    4                 0                 0                0                0
    5                 0                 0                0                0
    6                 0                 0                0                0
      sample id: DD172 sample id: DD176 sample id: DD188 sample id: DD216
    1                0                0                0                0
    2                0                0                0                0
    3                0                0                0                0
    4                1                0                0                0
    5                0                1                0                0
    6                0                0                1                0
      sample id: DD226 sample id: DD330 sample id: DD344 sample id: DD3516
    1                0                0                0                 0
    2                0                0                0                 0
    3                0                0                0                 0
    4                0                0                0                 0
    5                0                0                0                 0
    6                0                0                0                 0
      sample id: DD382 sample id: DD391 sample id: DD442 sample id: DD454
    1                0                0                0                0
    2                0                0                0                0
    3                0                0                0                0
    4                0                0                0                0
    5                0                0                0                0
    6                0                0                0                0
      sample id: DD456 sample id: DD474 sample id: DD506 sample id: DD513
    1                0                0                0                0
    2                0                0                0                0
    3                0                0                0                0
    4                0                0                0                0
    5                0                0                0                0
    6                0                0                0                0
      sample id: DD595 sample id: DD597 sample id: DD610 sample id: DD611
    1                0                0                0                0
    2                0                0                0                0
    3                0                0                0                0
    4                0                0                0                0
    5                0                0                0                0
    6                0                0                0                0
      sample id: DD6213 sample id: DD639 sample id: DD668 sample id: DD670
    1                 0                0                0                0
    2                 0                0                0                0
    3                 0                0                0                0
    4                 0                0                0                0
    5                 0                0                0                0
    6                 0                0                0                0
      sample id: DD672 sample id: DD69 sample id: DD712 sample id: DD714
    1                0               0                0                0
    2                0               0                0                0
    3                0               0                0                0
    4                0               0                0                0
    5                0               0                0                0
    6                0               0                0                0
      sample id: DD715 sample id: DD718 sample id: DD719 sample id: DD7533
    1                0                0                0                 0
    2                0                0                0                 0
    3                0                0                0                 0
    4                0                0                0                 0
    5                0                0                0                 0
    6                0                0                0                 0
      sample id: DD8379 sample id: DD88 sample id: ML138 sample id: ML161
    1                 0               0                0                0
    2                 0               0                0                0
    3                 0               0                0                0
    4                 0               0                0                0
    5                 0               0                0                0
    6                 0               0                0                0
      sample id: ML163 sample id: ML165 sample id: ML169 sample id: ML253
    1                0                0                0                0
    2                0                0                0                0
    3                0                0                0                0
    4                0                0                0                0
    5                0                0                0                0
    6                0                0                0                0
      sample id: ML257 sample id: ML283 sample id: ML292 sample id: ML298
    1                0                0                0                0
    2                0                0                0                0
    3                0                0                0                0
    4                0                0                0                0
    5                0                0                0                0
    6                0                0                0                0
      sample id: ML304 sample id: ML414 sample id: ML593 sample id: ML618
    1                0                0                0                0
    2                0                0                0                0
    3                0                0                0                0
    4                0                0                0                0
    5                0                0                0                0
    6                0                0                0                0


attr(,"assign")
  [1] 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
 [38] 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
 [75] 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
[112] 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
attr(,"contrasts")
attr(,"contrasts")$`factor(phenoData$characteristics_ch1.2)`
[1] "contr.treatment"

[bk11]

Your design in not correct. I suggest you to invest your time to learn this. Please follow the tutorial I have linked here which will help you understand how it is done-

https://sbc.shef.ac.uk/geo_tutorial/tutorial.nb.html

[bioinformatics]

                     DDLPS WDLPS
GSM766533     1     0
GSM766534     1     0
GSM766535     1     0
GSM766536     1     0
GSM766537     1     0
GSM766538     1     0
GSM766539     1     0
GSM766540     1     0
GSM766541     1     0
GSM766542     1     0
GSM766543     1     0
GSM766544     1     0
GSM766545     1     0
GSM766546     1     0
GSM766547     1     0
GSM766548     1     0
GSM766549     1     0
GSM766550     1     0
GSM766551     1     0
GSM766552     1     0
GSM766553     1     0
GSM766554     1     0
GSM766555     1     0
GSM766556     1     0
GSM766557     1     0
GSM766558     1     0
GSM766559     1     0
GSM766560     1     0
GSM766561     1     0
GSM766562     1     0
GSM766563     1     0
GSM766564     1     0
GSM766565     1     0
GSM766566     1     0
GSM766567     1     0
GSM766568     1     0
GSM766569     1     0
GSM766570     1     0
GSM766571     1     0
GSM766572     1     0
GSM766621     0     1
GSM766622     0     1
GSM766623     0     1
GSM766624     0     1
GSM766625     0     1
GSM766626     0     1
GSM766627     0     1
GSM766628     0     1
GSM766629     0     1
GSM766630     0     1
GSM766631     0     1
GSM766632     0     1
GSM766633     0     1
GSM766634     0     1
GSM766635     0     1
GSM766636     0     1
GSM766637     0     1
GSM766638     0     1
GSM766639     0     1
GSM766640     0     1
GSM766641     0     1
GSM766642     0     1
GSM766643     0     1
GSM766644     0     1
GSM766645     0     1
GSM766646     0     1
GSM766647     0     1
GSM766648     0     1
GSM766649     0     1
GSM766650     0     1
GSM766651     0     1
GSM766652     0     1
GSM766653     0     1
GSM766654     0     1
GSM766655     0     1
GSM766656     0     1
GSM766657     0     1
GSM766658     0     1
GSM766659     0     1
GSM766660     0     1
GSM766661     0     1
GSM766662     0     1
GSM766663     0     1
GSM766664     0     1
GSM766665     0     1
GSM766666     0     1
GSM766667     0     1
GSM766668     0     1
GSM766669     0     1
GSM766670     0     1
GSM766671     0     1
GSM766672     0     1
attr(,"assign")
[1] 1 1
attr(,"contrasts")
attr(,"contrasts")$group
[1] "contr.treatment"

Is this design better? Thanks!

[bk11]

It looks better than what you had before.