Dear all, would appreciate any suggestions on the following problem : how to correct for the changes in the basal level of gene expression, when comparing 2 conditions : let's say : a gene changes from 100 to 200 in response to hormone (fold change =2) the same gene changes from 150 to 250 in response to the same hormone, but a protein in the cell is knock-down in the experiment (fold change = 1.6) would need a procedure to correct for the changes in the basal level (100 to 150) in the experiments for a set of genes. any suggestions - perhaps using some scaling factors ? thanks ! b [[alternative HTML version deleted]]

On Fri, Oct 12, 2012 at 12:25 AM, Bogdan Tanasa <tanasa@gmail.com> wrote: > Dear all, > > would appreciate any suggestions on the following problem : how to correct > for the changes in the basal level of gene expression, when comparing 2 > conditions : let's say : > > a gene changes from 100 to 200 in response to hormone (fold change =2) > the same gene changes from 150 to 250 in response to the same hormone, but > a protein in the cell is knock-down in the experiment (fold change = 1.6) > > would need a procedure to correct for the changes in the basal level (100 > to 150) in the experiments for a set of genes. any suggestions - perhaps > using some scaling factors ? thanks ! > > Hi, Bogdan. What you are describing is a two-factor experiment. The factors, as I understand you, are treatment (with/without) and knockdown (with/without). If that is the case, then you should read up on using edgeR or DEseq on analyzing two-factor experiments. These methods naturally correct for both conditions and allow you to look for differences in gene expression due to treatment, knockdown, or (most likely what you are trying to find) both. Sean [[alternative HTML version deleted]]

On Fri, Oct 12, 2012 at 1:15 PM, Bogdan Tanasa <tanasa@gmail.com> wrote: > Thanks, Sean. It really helps. I would think it would be easy if I can > find an example that fits the case of our experiments (plus/minus > knockdown, plus/minus treatment) with no replicates. Thanks again ... > > Hi, Bogdan. To derive statistically interpretable results, you will want to get biological replicates. Both the edgeR user guide and the DEseq vignette include sections on experiments with multiple factors. You might want to work through those to get started. Sean > On Fri, Oct 12, 2012 at 3:55 AM, Sean Davis <sdavis2@mail.nih.gov> wrote: > >> >> >> On Fri, Oct 12, 2012 at 12:25 AM, Bogdan Tanasa <tanasa@gmail.com> wrote: >> >>> Dear all, >>> >>> would appreciate any suggestions on the following problem : how to >>> correct >>> for the changes in the basal level of gene expression, when comparing 2 >>> conditions : let's say : >>> >>> a gene changes from 100 to 200 in response to hormone (fold change =2) >>> the same gene changes from 150 to 250 in response to the same hormone, >>> but >>> a protein in the cell is knock-down in the experiment (fold change = 1.6) >>> >>> would need a procedure to correct for the changes in the basal level (100 >>> to 150) in the experiments for a set of genes. any suggestions - perhaps >>> using some scaling factors ? thanks ! >>> >>> >> Hi, Bogdan. >> >> What you are describing is a two-factor experiment. The factors, as I >> understand you, are treatment (with/without) and knockdown (with/without). >> If that is the case, then you should read up on using edgeR or DEseq on >> analyzing two-factor experiments. These methods naturally correct for both >> conditions and allow you to look for differences in gene expression due to >> treatment, knockdown, or (most likely what you are trying to find) both. >> >> Sean >> >> > > [[alternative HTML version deleted]]