Made a couple of tweaks to older files but the primary new stuff here is performing kruskall-wallis tests across the top 10% highest standard deviation LVs in all NF1 patient tumor samples, testing for significance as a function of the tumor type.
I went to start writing up the LV/multiplier results and this type of analysis seemed like the most logical starting point for that section. I also mapped the LVs to drug targets to help contextualize them a bit more when no pathway is associated with the LV.
Batch is pretty confounded with study here, with the exception of some MPNST samples, so there is not much we can do about that - worth keeping in the back of our mind and maybe mentioning as a limitation of these data in discussion.
allaway
commented
4 years ago
Made a couple of tweaks to older files but the primary new stuff here is performing kruskall-wallis tests across the top 10% highest standard deviation LVs in all NF1 patient tumor samples, testing for significance as a function of the tumor type.
I went to start writing up the LV/multiplier results and this type of analysis seemed like the most logical starting point for that section. I also mapped the LVs to drug targets to help contextualize them a bit more when no pathway is associated with the LV.
Batch is pretty confounded with study here, with the exception of some MPNST samples, so there is not much we can do about that - worth keeping in the back of our mind and maybe mentioning as a limitation of these data in discussion.