It is really common that the result of RNA-seq is different from other experiment for validation. There are always some uncertain factors that will effect your samples and make the results of same conditons divergent. Sometimes RT-qpcr can also get different results from RNA-seq, let alone enzyme assay that is on protein level. I have two suggestions:

1.Change the algorithm of calculating your different expression based on RNA-seq. There a lot of softwares that own diverse algorithm to count rpkm and their p-value:cufflinks, gfold, DEseq and DEG-seq.

2.Do more replicates to validate the most reliable difference.

Hello. The first thing that comes to mind in this case is the coverage (your sampling of the mRNA population in your cells) and the number of replicates you used for the experiment. If any of these are low you may have biased results. Validation with qRT-PCR may give you some hints on the quality of your sequencing.

It is really common that the result of RNA-seq is different from other experiment for validation. There are always some uncertain factors that will effect your samples and make the results of same conditons divergent. Sometimes RT-qpcr can also get different results from RNA-seq, let alone enzyme assay that is on protein level. I have two suggestions:

1.Change the algorithm of calculating your different expression based on RNA-seq. There a lot of softwares that own diverse algorithm to count rpkm and their p-value:cufflinks, gfold, DEseq and DEG-seq.

2.Do more replicates to validate the most reliable difference.

RNA-seq measures the amount of mRNA. Reporter gene enzyme assays measure the amount of protein made from that mRNA. The amount of protein depends on the amount of mRNA, but also on the amount of translation of the mRNA and the stability of the protein.

I agree with the technical problem in quantification as suggest in above answers however more important is, if your comparing quantification data both from RNA with protein expressed then you must take into consideration that not all the mRNA transcript can giverise to protein. Also mRNA stability is another parameter that may be going affect the comparison.

The stability of proteins and RNAs are different. Single point assays won't be very good for neither (especially for reporter proteins) -- transcriptional or translational rates are different at different stages of bacterial growth (lag, log and stationary). To get reproducible results, one has to sample multiple times during exponential phase of growth and do a differential plot -- activity vs. OD. The slope -- the differential rate -- should be almost linear thus the most reliable.

Good reporter products can be assumed to be stable, thus accumulative during growth. RNA turns over. It's probably worthwhile to compare the differential rate of a reporter with the RNA level. The former must be multiple points from a log phase -- say over a period of 2 doublings. The latter could be a single point -- as probably nobody does differential rate on RNA -- well into the log phase (say at least after 2 doublings after inoculation). But I'd still prefer to do two points for RNA, ~1 doubling apart, to be certain.