Entering edit mode
7.6 years ago
zizigolu
★
4.3k
hi,
I have two conditions A and B with 2 replicates. does it make any sense to use RPKMs for fold change or t-test analysis to identify differential expressed genes????
thank you
Is there a reason you don't want to use DESeq2/edgeR/limma?
you all right Devon,
I back home and I am with a lap top now... I should perform miRNA DE analysis..I have remote access to linux machine in MPI but it is too slow, I have fedora virtual box on my lap top but without anything installed such as python last night Ram helped me to equipped that but actually I am too naive in computer to be able to install python or like so, I tried galaxy but even uploading my files took many hours at last failed to complete... then my last chance is something working on windows like BBmap might be have at least RPKMs to perform t-test by stand alone tools. I don't have raw counts thereby no way to use featurecounts and then DESeq2 via R
If you have RPKM at gene level, i.e if they are "
not estimated counts
", you can simply convert them to raw counts using the RPKM formula.sorry you mean t-test between condition by RPKMs does not make sense?
sorry Devon,
I want to count the miRNA reads using the sorted .bam file containing only mapped reads (all generated by samtools from Bowtie output --SAM file) by featurecounts for Aspergillus fumigatus but in mirbase or wherever there was not a gtf for miRNA then I am using a GTF from ensembl. there is a gff3 contains miRNA but featurecounts did not recognize that. I converted that to gtf by cufflinks but again featurecounts did not recognize that. the GTF from enseble does not contain any miRNA then I am going wrong with this GTF??
thank you
You might need to write a little script to correctly convert the gff3 file to a standard GTF that other things can handle (GFF3 is really an annoying format).
thank you...scripting, my favorite!!! :)
when I converted gff3 to GTF by cufflink all miRNAs which were in gff3 disappeared in GTF