Total gene discovered/detected is less than gene in reference genome

pandapiggie · 08-04-2016, 01:20 AM

Dear All,

I did a RNA-seq on my bacteria (Salmonella) samples for a differential expression study. I used the cufflinks pipeline for this study.

When I checked the results of cuffdiff it showed 3752 genes but when I checked the Salmonella reference gene list it shows a total of 4949 genes. How is there such a massive difference in the gene numbers?

As I understand some of the genes may not be detected due to inactivation or are hypothetical genes. However this is a not satisfactory answer for me and I would like to know if anyone has an explanation for this. I may be looking at the wrong keywords so if possible can someone please point me in the correct direction for me to read up on this?

Thank you.

tristan dubos · 08-04-2016, 04:05 AM

Hi,
I think it's a good result ! I'm not specialist in bacteria but if i remember rigth there is a big diversity in population and the genomic dynamic is intensive ( i speak about bacteria in general i don t know if this rules match for Salmonella). But if I’m right if you add the genomic diversity and the temporal expression plus extraction bias, at the end you have 75% of the genes describes , is not bad , don't you think ?
Depends at the end what are you looking for ?
An other point is i don't know witch databases you use but you have to be care about it because you have 2 types of information :
-Gene describe (protein expressed function etc )
-Gene predicted , based on functional domain who are described on other protein and genomic structure/pattern

pandapiggie · 08-04-2016, 10:46 PM

Hi Tristan,

Thanks for the input. I totally did not take into consideration the RNA extraction bias and the sequencing bias as well.

Thanks for your help.

I was using the RefSeq complete genome reference from NCBI as my reference.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Annotate total variants against reference genome using CLC genomics?	Karlos	Illumina/Solexa	0	02-26-2015 08:13 AM
Digital gene expression in CLC without reference genome	RyNkA	De novo discovery	4	10-31-2013 07:14 PM
Difference between assembly, gene annotation, and reference genome?	prs321	Bioinformatics	4	08-29-2013 12:24 PM
entamoeba histolytica reference genome (gene annotation file) in GTF format	paula123	Bioinformatics	1	06-04-2012 03:39 AM

08-04-2016, 01:20 AM	#1
pandapiggie Junior Member Location: Malaysia Join Date: Jan 2016 Posts: 5	Total gene discovered/detected is less than gene in reference genome Dear All, I did a RNA-seq on my bacteria (Salmonella) samples for a differential expression study. I used the cufflinks pipeline for this study. When I checked the results of cuffdiff it showed 3752 genes but when I checked the Salmonella reference gene list it shows a total of 4949 genes. How is there such a massive difference in the gene numbers? As I understand some of the genes may not be detected due to inactivation or are hypothetical genes. However this is a not satisfactory answer for me and I would like to know if anyone has an explanation for this. I may be looking at the wrong keywords so if possible can someone please point me in the correct direction for me to read up on this? Thank you.

08-04-2016, 04:05 AM	#2
tristan dubos Member Location: France Join Date: Dec 2015 Posts: 39	Hi, I think it's a good result ! I'm not specialist in bacteria but if i remember rigth there is a big diversity in population and the genomic dynamic is intensive ( i speak about bacteria in general i don t know if this rules match for Salmonella). But if I’m right if you add the genomic diversity and the temporal expression plus extraction bias, at the end you have 75% of the genes describes , is not bad , don't you think ? Depends at the end what are you looking for ? An other point is i don't know witch databases you use but you have to be care about it because you have 2 types of information : -Gene describe (protein expressed function etc ) -Gene predicted , based on functional domain who are described on other protein and genomic structure/pattern

08-04-2016, 10:46 PM	#3
pandapiggie Junior Member Location: Malaysia Join Date: Jan 2016 Posts: 5	Hi Tristan, Thanks for the input. I totally did not take into consideration the RNA extraction bias and the sequencing bias as well. Thanks for your help. I was using the RefSeq complete genome reference from NCBI as my reference.