offTargets

@### For this:

//sequence dependent off-target effects (each separate shRNA vs. control, vs. the rest): So good shRNA at single copy vs Control (wt and mir30)? Each good shRNA at High copy vs Control? Is this what we want?//

It's either good or bad shRNA (at high or single copy) that we compare against the control (wt and miR30-emtpty) and against the other 5 shRNAs (at the same copy number - thus for high compare to the other 5 shRNA at high; for single compare to the other 5 shRNAs at single copy).

Then as a secondary measure of control, we can see how much overlap there is between the specifically deregulated genes for one shRNA at single and high copy.

Best, Christof

It is the checking for sequence specific off-target effects at high or low copy compared the the wt control

so this would take out off-target effects that appear with multiple sequences

all the probes from the array (grouped in different clusters / blocks) - the shRNAs that should be included are probably the 2 controls (wt, uninfected) as well as the 6 p53 shRNAs (I think we can exclude the Ren and Luci to make it more symmetrical with the high copy, wt control, single copy) - at the top the sequence-independent off-target effects (basically the plot you sent me earlier) - next, clusters of sequence-specific off-target effects for each of the 6 shRNAs (that basically would be a different color for only that shRNA at high copy; according to the numbers that should give a clear cluster for 3 shRNAs, a weak cluster for 1 shRNA and 2 shRNAs that have no off-target effect)

One of the follow up questions would then be: Why are there 3 shRNAs (1 potent, 2 weak) that do have sequence-specific off-target effects, while the other 3 don't? The second question would relate to the sequence independent off-target effects: Are there common miRNAs that we see deregulated - and due to what stage of processing - that can explain the collection of genes that is deregulated. I have small RNA seq of miRNAs for this, so that could serve for some model building.

c:Yes exactly! The idea here is to see if there are sequence specific off-target effect; thus they would have to be significantly different between the 6 p53 shRNAs, presumably at least at high copy. If that holds true, the next question is whether these effects are gone or not for the same shRNA at single copy.

v: My first question: You mention: Looking at one specific shRNAs (and doing that for all 6 p53 shRNAs) under wt & miR30-empty conditions vs. shX single vs. shX high: Meaning? Compare ShXSingle vs shX high is fine but what is the comparison between WT (wt and mir30 as replicates) and shX? Are you saying compare shX-high with WT and compare shX-low with Wt and then compare the two? (shX-high-WT) vs (shX-low-WT) ? If so I can do that. Just want to clarify.

The first, unbiased question would be if there are sequence dependent off-target effects, and whether we can avoid them solely by using the same sequences at single copy? Because this would obviously be the most relevant results as it is generally applicable.

Second, one could look into the sequence dependent off-targets that cannot be avoided with single copy, and see if they are mutually exclusive among different sequences.

Or, one could look into the shRNA-gene sequence relationship (as you mentioned) of the sequence dependent off-target effects that can be avoided with single copy, and try to figure out what type of hybridization pattern these shRNAs have (to figure out what is the minimum requirement etc.). However, the accuracy of the arrays might not be sufficient for that kind of an analysis.