StringTie's output can be processed by specialized software like Cuffdiff in StringTie's homepage.
However, the input files of cuffDiff include merged.gtf and bam files from all pairs of samples.
After all, the GTF file (Cd1h-1.gtf) output by Stringtie has one column for "TPM" value, but lost in the file t_data.ctab, the other output file of Stringtie.
What's more, we have no idea about Library normalization methods and
http://cole-trapnell-lab.github.io/cufflinks/cuffdiff/index.html#cross-replicate-dispersion-estimation-methods
Cuffdiff can use a global model for all conditions.
Should we use this when obtain the TPM from different samples and for further differential expression analysis?
So we post this mail for your suggestion on the parameter settings of CuffDiff to analyze the output of StringTie.
StringTie's output can be processed by specialized software like Cuffdiff in StringTie's homepage. However, the input files of cuffDiff include merged.gtf and bam files from all pairs of samples. After all, the GTF file (Cd1h-1.gtf) output by Stringtie has one column for "TPM" value, but lost in the file t_data.ctab, the other output file of Stringtie. What's more, we have no idea about Library normalization methods and http://cole-trapnell-lab.github.io/cufflinks/cuffdiff/index.html#cross-replicate-dispersion-estimation-methods Cuffdiff can use a global model for all conditions.
Should we use this when obtain the TPM from different samples and for further differential expression analysis? So we post this mail for your suggestion on the parameter settings of CuffDiff to analyze the output of StringTie.