I have 100 RNAseq files from 100 samples. using bioinformatics methods, I want to do final quality control and check if any technical problem in the data or not. since the files do not have annotation (treatment is not clear to me), I think PCA is not useful here since I do not know whether the samples that are clustered together have similar treatment or not (annotation is not clear). so the question is that what are the bioinformatics way to check if any technical problem happened (specially during sequencing)?
1 answer
First run fastqc, this can be parallelized e.g. with GNU parallel over all files depending on the amount of cores you have. Also quantify samples against the transcriptome, e.g. with a fast quasi-mapper such as salmon. You can combine the outputs of all tools into a summary report using multiqc. This will tell you quickly if something went wrong. If fastqc looks good and you good mapping rates everything is porbably ok from the technical side.
There are currently two ways of getting counts for genes in RNA-seq. Traditional alignment such as with STAR or hisat2 which align against the genome. Alternatively, there are now for a couple of years leightweight quantifiers that perform pseudoalignment or quasi-mapping. kallisto and salmon are prominent examples. They quantify reads directly against the transcriptome. Choice is yours, I personally use salmon.
Log in to answer this question.
-
losing many genes while doing DGE on the RNAseq data due to read count threshold
written by Sara
-
including incomplete data set in the PCA & gene expression analysis
written by Sara
-
novel AS events detection from RNA seq
written by Sara
-
low alignment rate of methyl-seq data
written by Sara
-
Strategies for merging RNAseq datasets
written by Nico80 •
-
CIBERSORT input file format
written by Sara
-
Quality Check Your Samples and Biological Replicates
written by eyonesi •
-
PCA on selected genes.
written by a.archana •
-
problem with the alignment of RNAseq data
written by Sara
-
problem with old UCSC browser links
written by Sara
