microRNA data analysis
0
0
Entering edit mode
6.2 years ago
illinois.ks ▴ 210

Hello,

While I am analysing microRNA data (bowtie2-htseq-count pipeline), I had a question! ( trim the adapter sequence, mapping with bowtie2) When, I aligned my sequence data, I have used the DEseq to call the DEGs.

I have two replicates for each of two comparision, which leads to the total of 4 samples. (After calling the htseq-count, as expected most of genes have 0 counts except microRNA data)

When I try to do the estimateDispersions function, it give warnings

"d <- estimateDispersions(d, method="blind", sharingMode="fit-only") " 
****glm.fit: algorithm did not converge****

Anyhow, I ran the analysis with

res = nbionomTest(d, "pre", "post")

When, I looked at the result, I found that weird observation.

I cannot see any significally expressed microRNA or anything.

    id  baseMean    baseMeanA   baseMeanB   foldChange  log2FoldChange  pval    padj
21074   TMEM107 78.02900309 25.72703226 130.3309739 5.065915594 2.34082304  0.101156951 1
19610   SNORD43 667.8657405 265.6019618 1070.129519 4.029072346 2.010447711 0.146159895 1
12696   MIR4521 1.600577469 3.201154937 0   0   #NAME?  0.151698991 1
22395   UTP3    1.600577469 3.201154937 0   0   #NAME?  0.151698991 1
12382   MIR3607 1.736586708 0   3.473173417 Inf Inf 0.172783874 1
19428   SNORD101    8.134565914 2.444304911 13.82482692 5.655933863 2.49976525  0.177692193 1
19609   SNORD42B    36.92888997 16.00577469 57.85200526 3.61444581  1.853774461 0.178053555 1

This is the top of my result. (the most significant gene has p value of 0.101156951), which means there is no significant genes. (is it because, in my htseq-count matrix, most of genes has zero counts???? )

Also, there are many of microRNA genes has large counts values, however none of them are identified as DEGs.

for example, the following microRNA p values are mostly around 0.8 ~ 0.9.

                pre           pre           post         post

MIR320A 8255    6799    10681   8168
MIR30A  757 629 924 673
MIR30B  1133    440 1620    335
MIR30C1 36  27  55  19
MIR30C2 31  25  55  21
MIR30D  19044   4202    40746   5721
MIR30E  1562    519 2953    546

However,

  pre pre post  post
MIR3607 0   0   4   4
MIR4521 2   3   0   0

these microRNA has more significant p value comparing with the upper microRNA, which seems to strange to me.

Does it make sense? Or do I make any mistakes?

RNA-Seq R • 1.3k views
ADD COMMENT
0
Entering edit mode

Are you actually using DESeq rather than DESeq2? If so, just stop what you're doing and switch.

ADD REPLY

Login before adding your answer.

Traffic: 2291 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