I want to do RNAseq on several blood samples for performing a differential expression analysis, specifically on a known pathway of around 100 genes.
I don't know if it's more relevant to do a classic whole mRNA transcriptome analysis, or should I ask for a custom panel to sequence only the 100 genes of the pathway.
In your opinion what would be the less expensive method?
Would a custom panel be relevant for differential expression, in term of normalisation of gene expression?
My guess is that standard RNA-seq is more suitable as it allows more sophisticated normalization. The established tools build on the assumption of most genes not changing for normalization, dispersion estimation etc. which is probably violated if you pull out like 100 genes. If you book a RNA-seq all-in service at a larger core facility (library prep + sequencing, you send in total RNA) we typically get offers of roughly 300-400€ / sample (HiSeq3000 single-end sequencing, for PE a bit more expensive). If you have the funds, I would definitely go for this. With these data you can go ahead with standard analysis without the need of digging out custom normalization strategies and differential analysis. I also guess that quality control and outlier detection, e.g. by PCA is difficult with only 100 genes which most likely will notably differ among your samples. As said, if you have the funds make your life easier and do RNA-seq.
Thank for you advice, I was indeed afraid of the normalisation with only 100 genes, I think I will go with whole transcriptome RNAseq. Would you know by any chance a good ngs provider in Europe?
We typically sequence at the Biomedical Sequencing Facility (BSF) in Vienna/Austria, though we make the libraries ourselves, but they offer all-in. Standard sequencing like a lane of a HiSeq3000 tyically is done within 3-4 weeks. Simply send a mail and ask for a quote. I would go for the standard polyA-enrichment workflow. https://www.biomedical-sequencing.org/
Still, you can also go for commercial companies like Novogene and others.
Thank for you advice, I was indeed afraid of the normalisation with only 100 genes, I think I will go with whole transcriptome RNAseq. Would you know by any chance a good ngs provider in Europe?
We typically sequence at the Biomedical Sequencing Facility (BSF) in Vienna/Austria, though we make the libraries ourselves, but they offer all-in. Standard sequencing like a lane of a HiSeq3000 tyically is done within 3-4 weeks. Simply send a mail and ask for a quote. I would go for the standard polyA-enrichment workflow. https://www.biomedical-sequencing.org/ Still, you can also go for commercial companies like Novogene and others.
Cool thank you very much for this information!