Hi Everyone,
I would like to ask for some advice about statistics for situations where we want to compare sequencing tag density profile alterations on genes, like in the case of some nascent RNA-seq studies.
Almost every sequencing techniques have a well established data analysis pipeline and corresponding statistics, i.e.: Gene expression (DEG) analyses, ChIP-seq for differential binding, etc. However, when I read about nascent RNA sequencing techniques (Gro-seq, Bru-seq, NETseq) that investigates Pol2 elongation rates or just different RNA biosynthesis profiles from ORFs, in some cases the used statistics are not straightforward or justified.
I.e.: lets say, I would like to carry out a similar experiment as it is in this paper:
I am using Bru-seq to measure transcriptional recovery upon DNA damage.
I have different time points after DNA damage (compared to control) and I generate tag density profiles for a selected gene in every time point after DNA damage.
In this case which statistics should be used to indicate that the recovery of transcription is significantly different on some gene, or the loss of RNA synthesis is significant or not?
Should I Bin the gene(s) and compare coverage per bin in each time point and use a "simple" paired T-test?
Once again, my question in a simplified way: if I want to compare RNA biosynthesis profiles / Pol2 binding profiles throughout gene bodies between conditions, what is the desired statistical method to calculate significance?
I would appreciate any help and advice.
I would like to ask for some advice about statistics for situations where we want to compare sequencing tag density profile alterations on genes, like in the case of some nascent RNA-seq studies.
Almost every sequencing techniques have a well established data analysis pipeline and corresponding statistics, i.e.: Gene expression (DEG) analyses, ChIP-seq for differential binding, etc. However, when I read about nascent RNA sequencing techniques (Gro-seq, Bru-seq, NETseq) that investigates Pol2 elongation rates or just different RNA biosynthesis profiles from ORFs, in some cases the used statistics are not straightforward or justified.
I.e.: lets say, I would like to carry out a similar experiment as it is in this paper:
I am using Bru-seq to measure transcriptional recovery upon DNA damage.
I have different time points after DNA damage (compared to control) and I generate tag density profiles for a selected gene in every time point after DNA damage.
In this case which statistics should be used to indicate that the recovery of transcription is significantly different on some gene, or the loss of RNA synthesis is significant or not?
Should I Bin the gene(s) and compare coverage per bin in each time point and use a "simple" paired T-test?
Once again, my question in a simplified way: if I want to compare RNA biosynthesis profiles / Pol2 binding profiles throughout gene bodies between conditions, what is the desired statistical method to calculate significance?
I would appreciate any help and advice.