QQ and density plots with one replicate in EBseq HMM
54 views
Skip to first unread message
Diana Duarte
Apr 8, 2016, 11:11:00 AM4/8/16
to EBSeq users
Hi
I have an experiment in 5 time points without replicates with around 50 000 transcripts. I filter the data base and kept the transcripts with at least two reads in each time point. I ran the data with 10 and 20 iterations and the LL values do not change that much between them. I will attach the resulting diagnostic plots with 10 iterations. I do not have many experience interpreting these graphics. The first time point do not show such a good fit according to qq plot for me... and in the histogram i see this also for the first time point... Is there a way to improve the fit in this case for this time point? For me the other histograms do not show such a good fith either... or aren they that bad? The qq plots for the other time points for me look in general ok... but as i told you, i do not have many experience interpreting these graphics. Any advice is highly appreciated.
I have an experiment in 5 time points without replicates with around 50 000 transcripts. I filter the data base and kept the transcripts with at least two reads in each time point. I ran the data with 10 and 20 iterations and the LL values do not change that much between them. I will attach the resulting diagnostic plots with 10 iterations. I do not have many experience interpreting these graphics. The first time point do not show such a good fit according to qq plot for me... and in the histogram i see this also for the first time point... Is there a way to improve the fit in this case for this time point? For me the other histograms do not show such a good fith either... or aren they that bad? The qq plots for the other time points for me look in general ok... but as i told you, i do not have many experience interpreting these graphics. Any advice is highly appreciated.
Reply all
Reply to author
Forward
0 new messages