SEQanswers

Go Back   SEQanswers > Bioinformatics > Bioinformatics



Similar Threads
Thread Thread Starter Forum Replies Last Post
Pan-Cancer lncRNA, RNA-RNA and Protein-RNA networks from TCGA 14 cancer types yjhua2110 Bioinformatics 13 05-18-2015 03:49 AM
DEseq2: can filtering low count genes reduce the number of DE genes? JonB Bioinformatics 2 06-10-2014 06:15 AM
Protein Coding genes gokhulkrishnakilaru Bioinformatics 0 02-10-2012 10:10 AM

Reply
 
Thread Tools
Old 05-06-2015, 11:23 AM   #1
stormin
Member
 
Location: US

Join Date: Aug 2014
Posts: 23
Default What types of genes/protein are lost with low depth?

Methods for both determining differential expression of mRNA and protein expression levels depend on the depth of the sequencing/MS runs. Are there any previous studies that systematically looked at what types of mRNA/protein are low expressing and thus will be lost with low depth? For example, are kinases typically low expressing proteins and will be lost without high depth?
stormin is offline   Reply With Quote
Old 05-06-2015, 11:57 AM   #2
Richard Finney
Senior Member
 
Location: bethesda

Join Date: Feb 2009
Posts: 700
Default

It's pretty much basic statistics, isn't it?
One uniquely mapped read is evidence of transcription.
Richard Finney is offline   Reply With Quote
Old 05-06-2015, 08:52 PM   #3
stormin
Member
 
Location: US

Join Date: Aug 2014
Posts: 23
Default

Actually I mean if there are any prior research in what types of protein/genes are typically low expressing. For example, we know housekeeping genes are typically always on and expressed in cell at fairly high levels. What about the lower end?
stormin is offline   Reply With Quote
Old 05-06-2015, 10:37 PM   #4
Melissa
Senior Member
 
Location: Switzerland

Join Date: Aug 2008
Posts: 124
Default

Hmm... I'm curious if there is such a list. A lowly expressed gene is either not activated or repressed. I think some people are overly concern about not getting sufficient depth of lowly expressed genes. The depth of the sequencing doesn't change the fact that a gene is lowly expressed but it can provide a more reliable read count across replicates. If a gene is lowly expressed in all your samples, it should be removed from analysis because it will not give statistically significant results in differential gene expression.
Melissa is offline   Reply With Quote
Old 05-07-2015, 07:03 AM   #5
Richard Finney
Senior Member
 
Location: bethesda

Join Date: Feb 2009
Posts: 700
Default

"A lowly expressed gene is either not activated or repressed" ... or ... doesn't need a whole lot of expression to do it's job.
I'm not sure setting a low number of reads to zero solves anything.
Richard Finney is offline   Reply With Quote
Old 05-07-2015, 08:39 AM   #6
mbblack
Senior Member
 
Location: Research Triangle Park, NC

Join Date: Aug 2009
Posts: 245
Default

Quote:
Originally Posted by stormin View Post
Actually I mean if there are any prior research in what types of protein/genes are typically low expressing. For example, we know housekeeping genes are typically always on and expressed in cell at fairly high levels. What about the lower end?
Well, there are no such things as truly universal housekeeping genes. For any gene that one person has found useful as a housekeeping gene, you can almost certianly find an example where it is differentially expressed or not expressed. Housekeeping genes are context dependent.

Any list of "low expressors" is pointless without identifying species, tissue, life stage and environmental condition at the time of sampling. All you have to do is look at what is known about tissue specific expression profiles to realize there is no inherently universal such gene or genes.
__________________
Michael Black, Ph.D.
ScitoVation LLC. RTP, N.C.
mbblack is offline   Reply With Quote
Old 05-07-2015, 09:05 AM   #7
stormin
Member
 
Location: US

Join Date: Aug 2014
Posts: 23
Default

Quote:
Originally Posted by mbblack View Post
Well, there are no such things as truly universal housekeeping genes. For any gene that one person has found useful as a housekeeping gene, you can almost certianly find an example where it is differentially expressed or not expressed. Housekeeping genes are context dependent.

Any list of "low expressors" is pointless without identifying species, tissue, life stage and environmental condition at the time of sampling. All you have to do is look at what is known about tissue specific expression profiles to realize there is no inherently universal such gene or genes.
Thanks for all of the responses. To me, this sounds like it's coming from personal experience of looking at multiple tissue specific profiles. I am wondering if any prior papers have delve into this topic and have reached a quantitative answer on this.
stormin is offline   Reply With Quote
Old 05-07-2015, 09:56 AM   #8
mbblack
Senior Member
 
Location: Research Triangle Park, NC

Join Date: Aug 2009
Posts: 245
Default

Yes, I've done some comparative studies of multiple tissues from the same animals (eg. rodent liver & lung from inhalation studies) and one can get near completely opposite differential expression of the same genes between tissues. Even many kinases can be up or down regulated under some chemical exposures, and may be differentially affected in different tissues. After 20 odd years of studying gene expression in invertebrates and mammals, I'm not at all surprized any more when any particular gene is highly differentially expressed (even when I admit I cannot make sense of some of them given the apparent biological context of the experimental condition I'm studying).

Expression is highly context dependent and there is a lot of literature on that. Species, sex, life stage, tissue, nutritional state, health state, environmental affects, inter-population differences, etc, etc - all of these things are known to impact measured gene expression. Given that, I'll play devils advocate and ask why you would even expect that there would be such a thing as an invariantly low expression gene in the first place?

And, keep in mind that RNA-seq is not an absolute quantification of expression to begin with. It is a relative estimate of expression and we know that not all available RNAs in a cell are sampled equally in any sequencing experiment (which also in turn then begs the question of how low is considered low expression - what's your cutoff there, and how variable is invariant?). Even were you to control for as many variables as possible, but did a sampling time course every 30-60 minutes for a day (or whatever), how much variation do you think you would see in many of your low count features just by sampling the same pool of cells/tissue across a diurnal cycle (or a feeding cycle, or a menstrual cycle or any other biologically cyclical process)?
__________________
Michael Black, Ph.D.
ScitoVation LLC. RTP, N.C.

Last edited by mbblack; 05-07-2015 at 10:07 AM.
mbblack is offline   Reply With Quote
Old 05-07-2015, 11:07 AM   #9
stormin
Member
 
Location: US

Join Date: Aug 2014
Posts: 23
Default

Hi mbblack, I think expression level can be correlated with function and particular proteins, due to their function, may have very low gene expression level and/or protein expression level because they are very potent in their function.

I do agree that different situations will up or down-regulate these genes/protein, but I am asking in general are there any studies that looked at whether any particular groups that fit this profile.
stormin is offline   Reply With Quote
Old 05-08-2015, 05:01 AM   #10
mbblack
Senior Member
 
Location: Research Triangle Park, NC

Join Date: Aug 2009
Posts: 245
Default

Perhaps average absolute expression levels for a given wild type tissue may behave that way. But you would not be able to address that question by sequencing whole cell mRNA.

Honestly, I'd be more inclined to think that perhaps mean constituent cellular protein concentrations may behave that way for a given protein in a given tissue or cell type, but not necessarily expression of the genes for that protein. To my thinking, if a relatively invariant constituent protein level is necessary for some functional or regulatory aspect, then expression may likely have to be very labile for that protein in order to maintain relatively constant cellular protein levels over time.

P.S. this paper may appear a bit dated, but its message is still highly relevant to this sort of discussion - it is for yeast: http://citeseerx.ist.psu.edu/viewdoc...=rep1&type=pdf
__________________
Michael Black, Ph.D.
ScitoVation LLC. RTP, N.C.

Last edited by mbblack; 05-08-2015 at 05:22 AM.
mbblack is offline   Reply With Quote
Reply

Thread Tools

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off




All times are GMT -8. The time now is 07:35 PM.


Powered by vBulletin® Version 3.8.9
Copyright ©2000 - 2020, vBulletin Solutions, Inc.
Single Sign On provided by vBSSO