Dear Thushara,

There are few things about the two techniques which make them bit different.

1. RT-qPCR is more sensitive than RNA-seq as you employ gene specific primers to detect your gene of interest from the cDNA whereas RNA-seq targets a whole population.

2. RT-qPCR: you normalise the measurements to a house keeping gene to normalise for the amount of RNA/cDNA in your sample but in RNAseq the normlaization is for the population/library size etc.

3. Depends on the sequencing depth (coverage) you may or may not quantify all transcripts of a gene in your sample.

4. Depends on whether the genes A, B and C have paralogues in the genome or partial matches some reads may not counted as aligned by the aligner you used.

What you may do:

1. Make sure the quality of your RT-qPCR data is good. No primer dimers, NTC controls are clean etc. and you used the correct house keeping gene for the condition (The house keeping gene transcript levels are not much altered between control vs treatment etc.)

2. Look at the data in a genome viewer such as IGV. You can load indexed BAM files and see how the reads look. Is there a huge variation between replicates etc.

3. Check whether the genes have duplicates/ shared regions in the genome and the read alignment settings allow multiple alignments etc.

Goodluck.

Shajahan

RNA seq reads may also include many variants, splicing or differential promoters or poly A sites for the genes that you are not even considering or may not know about. Your PCR only detects one specific region that your primers recognize.

Thank you so much for your answers!!

We used GAPDH as HK whose readings were very tight in our samples.

I'm more inclined to think about alternatively spliced transcripts as these proteins are not well studied and is thought to behave differently in different cell types. If that is the case, which data is more reliable? Isn't it RNA seq since it tells you about the entire transcriptome?

Unfortunately, you can't say so. If your qPCR primer pair amplifies an exon common to all isoforms then your qPCR measures the sum of all isoforms. You can predict splice variants based on the sequence. Depends on the organism there could be expression data to suggest if that's observed. IGV alignments of bam files may suggest an isoform bias if there is alternative splicing happening in A, B or C and one isoform differentially dominates in your treatment. You can also simply test this hypothesis by running a qPCR with transcript specific primers unique to one isoform if there the predicted isoforms have unique exons.

Thank you for your inputs!! we plan to do follow up QPCR to confirm.