Thursday, January 20, 2011

NanoString Tryout

You might remember that previously I reported on our efforts on RNA extraction from yeast cells. Since then Assaf and Ayelet quantified the amounts and quality of RNA we get by doing RNA purification from the cell supernatant in Assaf's assay. They showed that the amount of RNA we get matches what literature reporters (about 1.2 pico-gram/cell), and that the RNA does not show any degradation.

Last week Assaf performed a time course experiment where he grew yeast, and then subjected them to salt stress, and collected cell at different time points following the stimulation. He froze samples of cells by droping them into very cold (< -40C) methanol, and then extracted supernatant using the technique I discussed in the previous entry. 

He packed the supernatant in a box with dry ice and we sent it to Ollie's lab at U. Mass. Medical School by carrier service. We were not sure if the samples will get there in good shape, and it was great to hear that they did, still frozen. Hsiuyi from Ollie's lab run the samples on the Nanostring nCounter machine they have there and sent us back the results.

Few words about the NanoString machine (you can read a nice overview on their website). The basic idea is that we order a set of probes that match genes we are interested in, and then using these probes can count individual molecules in a small volume sample. By counting how many times we see each type of probes we get a very accurate measure of the number of RNA molecules in the initial sample. The cool thing about this method is that it does not use any enzymatic reaction and thus does not have biases due to PCR preferences and such. Moreover, it is very simple to run and does not require RNA purification, labeling and such.

We were very happy that the results we got from Hsiuyi showed high degree of reproducability. Two repeats of the experiments looked almost identical (if they would have been more similar I would start being suspicious).


In this figure each dot correspond to a gene, and the x/y axis are the log (base 2) of the number of counts we had for that gene in the first or second replicate expriment. We can see that the dynamic range here is between 2 to 16, so we have 2^14 (= 4096) fold difference between the lowest probe and the highest one.

This success shows that Assaf's setup is ready for the big experiment of measuring many different mutants. Hopefully we will have this done in the next few weeks.

1 comment:

Anonymous said...

Would you by any chance remember the correlation (more or less than 0.9) ? Also, were/are there any red flags regarding the high reproducibility ?