Cuffdiff output

[sbdk82]

I am running BWA->cufflink for differential gene expression and the cuffdiff output contains many repeated gene_ids. For example:

test_id	 gene_id	 gene	locus	sample_1 	 sample_2 	 status 	value_1	value_2	log2(fold_change)	 test_stat	 p_value	 q_value	 significant
XLOC_000665	XLOC_000665	Traes_1AL_34404D5D8	IWGSC_CSS_1AL_scaff_3881073:2596-3747	q3	q5	OK	42.1649	2.81737	-3.90362	-8.49777	5.00E-05	0.00921445	yes
XLOC_000665	XLOC_000665	Traes_1AL_34404D5D8	IWGSC_CSS_1AL_scaff_3881073:2596-3747	q4	q5	OK	16.8182	2.81737	-2.5776	-5.45754	0.00025	0.0386488	yes

Can anyone explain why there are repeats like this?

Also, I need to use BLAST against similar plant species to find information about the genes which are differentially expressed. I am not able to understand the "locus" column. In the above example, the locus is IWGSC_CSS_1AL_scaff_3881073:2596-3747. In the reference file there is a sequence associated with IWGSC_CSS_1AL_scaff_3881073. Do I need to select the subsequence starting from location 2596 to 3747 for BLAST?

[yehudithasin]

Hi sbdk.

Cuffdiff outputs a line for each ID, for each comparison.
In the example you posted the first line is comparison between q3 and q5, and the second is between q4 and q5 for the same gene.
Yu

[sbdk82]

Thanks Yu !!

Actually I have two samples of data with three replicates in each sample. Thus I have 6 fasta files. So each data file represented by qi (i = 1,2 ..6) ? I used cufflink and then cuffmerge before using cuffdiff.