FAQ  •  Register  •  Login

Monte-Carlo Simulation

Forum rules
Please be as geeky as possible. Reference, reference, reference.
Also, please note that this is a mixed bag of math-gurus and mathematically challenged, so choose your words wisely :-)



Posts: 1

Joined: Fri Oct 20, 2017 9:39 pm

Post Fri Oct 20, 2017 10:41 pm

Monte-Carlo Simulation

I was wondering if anyone had tried this:
Generate an array of random values, with the x axis being markers of interest and the y axis being individual events. Optimally, the range of values would have a mean and standard deviation equal to the marker in that row. Then run this array through Phenograph and see what sort of clustering it does.

Has anyone else tried running an array of random values through a clustering algorithm.



Posts: 2162

Joined: Fri Nov 01, 2013 5:30 pm

Location: Stanford HIMC, CA, USA

Post Mon Oct 23, 2017 2:42 pm

Re: Monte-Carlo Simulation

I am not aware of anyone having done this as a simulation for CyTOF data.

However, I think there's been more work done on this for scRNASeq and other genomics work......you might search for papers in that area. Many of the flow and scRNASeq algorithms are related, so it might be informative.



Posts: 22

Joined: Thu Nov 19, 2015 4:23 pm

Post Mon Oct 23, 2017 2:48 pm

Re: Monte-Carlo Simulation

Sorry I didn't get here sooner. Aaron Lun, a well regarded RNA seq analysis researcher put out a very nice package cydar http://bioconductor.org/packages/release/bioc/vignettes/cydar/inst/doc/cydar.html that has a section for generating random data. If you were to run random data through any clustering algorithm, you would still get clusters. You would still expect clusters to show evidence of differential abundance and expression as well. However, once you adjusted your p values for all those comparisons, you would expect to observe a distribution of p values that were flat. Hope that helps.


Return to CyTOF data analysis

Who is online

Users browsing this forum: No registered users and 1 guest