Don't ask to ask, just ask | |
PerlMonks |
comment on |
( [id://3333]=superdoc: print w/replies, xml ) | Need Help?? |
@bliako - thank you for your investigation. I think I misspoke when I said "there is no theoretical or empirical indication that it would make a difference while assessing background signal from these shuffled sequences" What should I have instead said is that "periodicities and abundance biases of di-, tri-, or multi-mer sequences when completely destroyed are more likely to be truly random shuffles, rather than shuffles that retain frequencies of di-mer, or tri-mer etc. This is obvious from your own empirical observation from one human chromosome Anyways, I have an old script that uses shuffle from List::Util Perl module and I remember it never ran out of RAM, so I think I will use that. Though I would have preferred something that concats the entirety of the genome and then shuffles and fragments as per length distribution of chromosomes in input - but various Perl scripts that do this run out of RAM even on my cluster. Need to take a weekend break to figure out RAM usage and fix it. In reply to Re^3: Reduce RAM required
by onlyIDleft
|
|