Cuffdiff output file gene_exp.diff has "no" for all listed genes under "significant" head.
0
0
Entering edit mode
4.3 years ago

In context to the aforementioned statement, is that to say that my analysis is egregious hitherto.

Cuffdiff RNA-Seq • 808 views
ADD COMMENT
0
Entering edit mode

In cuffdiff significant value (yes or no) defined by q_value cut off. Do you have replicates ? if you don't have replicates there is no meaning of p or q value.

ADD REPLY
0
Entering edit mode

You mean it would therefore be pointless to analyze data using this pipeline if I don't have replicates, or will the data not be very reliable? And if the data cannot be analyzed without replicates, would it be OK for me to use the different fastq files (paired-end reads) as replicates, as both the fragments would map to the same reference genome (I understand that not all paired-end reads will have fragments mapping to the same gene)?

ADD REPLY
1
Entering edit mode

Hi,

I am not telling that it is pointless to use cuffdiff if you don't have replicates. If you don't have replicates there is no meaning of p value or q value. Read this to know more about p value and how to calculate it for diff expressed genes. You have to be more stringent to get your differentially expressed genes if you don't have replicates. (for e.g fold change >= 2 , minimum fpkm/rpkm 10 or some arbitrary cutoff on number of reads mapped to particular gene.)

Replicates are there to check if biology can be reproduced or not. By using r1 and r2 as replicates (which are not replicates in any sense) you are not checking biology but something else (which I don't know)

ADD REPLY
0
Entering edit mode

You need to provide details such as command you used, and things you tried so far.

ADD REPLY

Login before adding your answer.

Traffic: 1465 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