Coverage function to turn df into Rle?

[endrebak]

The original IRanges library has a function called coverage to turn interval data into an Rle. Will you consider implementing a coverage function?

Thanks for the library, btw.

[phaverty]

Hi Endre, I'm glad you are benefitting from the package. My GenomicVectors.jl package has a coverage method that returns an RleVector. I think there is room for optimization. Please let me know how it works for you. Pete

…

____________________ Peter M. Haverty, Ph.D. Genentech, Inc. phaverty@gene.com

On Wed, Apr 11, 2018 at 9:29 PM, Endre Bakken Stovner < ***@***.***> wrote: The original IRanges library has a function called coverage to turn interval data into an Rle. Will you consider implementing a coverage function? Thanks for the library, btw. — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#12>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AH02K6h1B5_wKiyVrSE34aE5OfktBm0Iks5tntg7gaJpZM4TRH7X> .

[endrebak]

Thanks. I searched the wrong repo then.

My intention is to translate it into Cython - I'll be sure to give you credit. I have an implementation myself, but it uses two heaps.

[endrebak]

Julia looks nice! I need to learn it to be able to understand how you can write a coverage-function as simply as this, because I am guessing you are not instantiating a vector of chromosome length:

function coverage(gr::AbstractGenomicVector)
    out = RLEVector(0, last(chr_ends(gr)))
    for (s,e) in eachrange(gr)
        out[s:e] += 1
    end
    out
end

[phaverty]

Hi Endre, The subtypes of AbstractGenomicVector contain a GenomeInfo object, which contains the chromosome lengths (actually, the cumsum of the lengths). This allows AbstractGenomicVector to store the genome base positions as the offset from the first base of chr1. That way we can do (almost) everything on one scale without worrying about chromosomes. And since julia has 64bit ints we can have one RLE that spans the genome. (R only has 32 bit, signed integers, which are to small to hold an index into the last few chromosomes). Pete

…

____________________ Peter M. Haverty, Ph.D. Genentech, Inc. phaverty@gene.com

On Fri, Apr 13, 2018 at 4:55 AM, Endre Bakken Stovner < ***@***.***> wrote: Julia looks nice! I need to learn it to be able to understand how you can write a coverage-function as simply as this, because I am guessing you are not instantiating a vector of chromosome length: function coverage(gr::AbstractGenomicVector) out = RLEVector(0, last(chr_ends(gr))) for (s,e) in eachrange(gr) out[s:e] += 1 end out end — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#12 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AH02K3B4VVEC0iIJt-DS90Xz_KQdcVInks5toJIigaJpZM4TRH7X> .