Hi Sam, Thank you for your reply! My only concern is my cells stopped growing / slow down after the treatment, so that the cell densities at certain time point are different between control and treated cells, while the time density at time points can be similar as Time zero. Thus I am worried some pathways that are controlled related to cell density will be transcribed differently. So should I have two controls: "a time zero no treatment" and "time point no treatment"? Thank you for your opinion! Best, Nancy On Fri, Jun 14, 2013 at 11:47 PM, Sam McInturf <smcinturf@gmail.com> wrote: > Nancy, > It depends on what questions you would like to ask. If you would like > to ask "what genes are respond to treatment at time 1" and then "what genes > respond to treatment at time 2" and so on, then you need to run a control > sample for each time point. If you only have untreated at time 1, then you > can as "what genes respond differently in time 1 untreated compared to time > 0 untreated. Although, it is a hard argument to make that the global gene > expression is constant over time, so this comparison is unlikely to be > biologically meaningful. > > If you are unfamiliar with linear models and other statistics topics and > you do not intend on learning them before you do your experiment, contact a > bioinformatician and let them design the experiment, it will probably save > you trouble in the long run. > > So, to directly answer your question, you don't HAVE to run a control at > each time, but you will not be able to answer many meaningful biological > questions. RNA sequencing is currently very expensive, but if you skimp > out on replicates / time points/ et cetera, you are just going to waste > money, > > Best > > > On Fri, Jun 14, 2013 at 12:44 AM, Yanxiang Shi <nancyyxshi@gmail.com>wrote: > >> Hi, >> >> I am planning to do RNA-seq on bacterial samples collected at different >> time points after treatment. I am just wondering, what should be the >> control of this series of samples, should it be untreated sample at time 0 >> right before the treatment, or should it be untreated samples at >> corresponding time points? As the RNA-seq cost a good amount of money, I >> think "untreated samples at different corresponding time points" would >> cost >> a lot, but which way would make more sense both biologically and >> statistically? I will do triplicates for each time point and condition. >> >> I am new to this field so get really confused some time. Thank you very >> much in advance!!! >> >> Best, >> Nancy >> >> [[alternative HTML version deleted]] >> >> _______________________________________________ >> Bioconductor mailing list >> Bioconductor@r-project.org >> https://stat.ethz.ch/mailman/listinfo/bioconductor >> Search the archives: >> http://news.gmane.org/gmane.science.biology.informatics.conductor >> > > > > -- > Sam McInturf > [[alternative HTML version deleted]]