Why is mappability removed?

[magicDGS]

One of my main usages of gem-mapper is to compute a mappability track, but I just found that it is removed in the latest commit to master. Is there any plan to include it again in a follow-up commit? Otherwise, is there any other software to compute the same mappabiltiy track as gem?

Thank you!

[smarco]

Daniel,

The mappability was temporally removed as it was producing slightly different results compared to the old mappability. For that reason, we wanted to look in detail to those small differences.

And yes, we will make it available again soon. Meanwhile, you can use the old version of GEM to compute the mappability.

Cheers,

On Tue, May 1, 2018 at 11:50 AM, Daniel Gómez-Sánchez < notifications@github.com> wrote:

One of my main usages of gem-mapper is to compute a mappability track, but I just found that it is removed in the latest commit to master. Is there any plan to include it again in a follow-up commit? Otherwise, is there any other software to compute the same mappabiltiy track as gem?

Thank you!

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/smarco/gem3-mapper/issues/7, or mute the thread https://github.com/notifications/unsubscribe-auth/ACdvbUp5XnCTaOVJ5aBDGvSZwf7POpPNks5tuC_mgaJpZM4Ttwv6 .

--

Santiago Marco-Sola

[magicDGS]

Thanks for the quick answer, Santiago. I tried to find the old version of mappability at http://gemlibrary.sourceforge.net, but it redirects to http://dat.cnag.cat/wiki/The_GEM_library, which is returning a 404 response. Thus, I cannot access the old binaries anymore - do you have any clue why the page is not longer available?

[smarco]

Daniel,

I'm very sorry, this is very unfortunate. Paolo Ribeca, the owner of the web page, has already been noticed and he'll fix it ASAP.

Let me point you to sourceforge directly: https://sourceforge.net/projects/gemlibrary/files/gem-library/

Let me know. Best,

On Wed, May 2, 2018 at 7:57 AM, Daniel Gómez-Sánchez < notifications@github.com> wrote:

Thanks for the quick answer, Santiago. I tried to find the old version of mappability at http://gemlibrary.sourceforge.net, but it redirects to http://dat.cnag.cat/wiki/The_GEM_library, which is returning a 404 response. Thus, I cannot access the old binaries anymore - do you have any clue why the page is not longer available?

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/smarco/gem3-mapper/issues/7#issuecomment-385872494, or mute the thread https://github.com/notifications/unsubscribe-auth/ACdvbdAgfFE_26LZoiF9UFNkIkSO-d4fks5tuUrJgaJpZM4Ttwv6 .

--

Santiago Marco-Sola

[magicDGS]

Thanks!

[magicDGS]

Any news about the mappability tool?

[smarco]

It's our intention to release soon a fully functional version within the next update of the tool.

For the time being, you can use the old mappability tool at https://sourceforge.net/projects/gemlibrary/files/gem-library/

Sorry for the inconvenience,

[lldelisle]

Hi, I would need the gem-mappability in galaxy. If you are close to update I would prefer to wait and use the last version. Could you tell me if you have an idea of the date? Thank you very much,

Lucille

[smarco]

Hi Lucille,

Thought I'm actively working on it, I cannot guarantee that the new mappability will be released before a couple of months. I'm really sorry about any inconvenience it can cause.

Please, use the current mappability tool for the moment (https://sourceforge.net/projects/gemlibrary/files/gem-library/).

Thanks for using GEM,

[lldelisle]

Many thanks for the reply.

[splaisan]

Also an old and enthusiast user of mappability and adding my vote here to get the new version as soon as you can make it. Thanks in advance Stephane

[duartemolha]

Hi ... I followed the description here https://wiki.bits.vib.be/index.php/Create_a_mappability_track

but I had created the index using the latest version of gem-indexer

gem-mappability complains about the header of the index file

is the index file for the latest version of gem3-mapper incompatible with the archive version?

Thanks

[splaisan]

Hi @duartemolha, please read [https://github.com/smarco/gem3-mapper/issues/10](this other note) to solve the issue

[duartemolha]

Hi @duartemolha, please read [https://github.com/[/issues/10](https://github.com/smarco/gem3-mapper/issues/10)](this other note) to solve the issue

Thanks but I believe the problem I encountered is before the gem-2-wig step... I am failing to produce the mappability file ...

create mappability track

This step :

gem-mappability -T 8 -I hg38_index.gem -l 150 -o hg38_index_150

[smarco]

Hi @duartemolha,

Please make sure you are generating the index using the same version as the mappability (https://sourceforge.net/projects/gemlibrary/files/gem-library/). Note that GEM3 indexes are not backwards compatible.

[duartemolha]

thanks

I understand now

[duartemolha]

any update on when mappability might be incorporated back into the software?

[VidJa]

May I vote for the mappability option as well?

[duartemolha]

BTW... I am creating a mappability at 150bp for the latest human reference hg38. Do you know if someone has already done that? I am running it in a pretty powerful server and I estimate it will take at least 24 hours

[smarco]

Honestly, I cannot commit to any specific deadline, but I can assure you it's among my top priorities.

[speromelior]

Any news about the mappability feature in the current mapper? Is this the only way to get a mappability track, using the old software? https://evodify.com/gem-mappability/

[VidJa]

Im experimenting with GenMap

https://www.biorxiv.org/content/10.1101/611160v1

Op di 10 dec. 2019 om 14:46 schreef speromelior notifications@github.com

Any news about the mappability feature in the current mapper? Is this the only way to get a mappability track, using the old software? https://evodify.com/gem-mappability/

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/smarco/gem3-mapper/issues/7?email_source=notifications&email_token=ABOTH3UKGA7I3LBMKPRCEE3QX6MUXA5CNFSM4E5XBP5KYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEGPI4MY#issuecomment-564039219, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABOTH3SKFBM7RYEEFYQR45DQX6MUXANCNFSM4E5XBP5A .

[smarco]

@speromelior I'm afraid not. For the moment, the old gem-mappabilty seems to be working and returning good results (https://evodify.com/gem-mappability/). We will inform you if we finally port it to GEM3.

@VidJa thanks for your update. Let us know about your experience.

[splaisan]

Hi again, still no good news I am afraid. Today I tried to compute mappability for long reads after choosing 8kb as kmer length. This is apparently too much asking to the gem software since I was still at 0.3% of my data after 24h. Is there another method I could apply to compute a mappability track for a very repetitive plant genome using long kmers in the kb range (eg pacbio long-reads) One additional issue is very likely the platform error rate (CLR reads, not CCS) here which makes this even more difficult. Can someone comment on the idea and maybe suggest tools? Thanks

[speromelior]

I couldn’t get gem3 to work for hexaploid wheat and I am studying a large gene family for which a track showing unique regions in the reference is very useful. My use was to compute the mappability of a reference rather than long reads (not sure about mappability of reads prior to assembly…) but for reference mappability I found genMap (https://github.com/cpockrandt/genmap) worked well, a utility based on the robust SeqAn libraries. I also found the GenMap author cpockrandt very helpful; the reference index needs a lot of memory and temp space to construct, I got there in the end, but before I did the author had posted one on his site.

Good luck and best regards,

Dan Smith

From: Stephane Plaisance notifications@github.com Sent: 04 June 2020 13:54 To: smarco/gem3-mapper gem3-mapper@noreply.github.com Cc: Dan Smith dan.smith@rothamsted.ac.uk; Mention mention@noreply.github.com Subject: Re: [smarco/gem3-mapper] Why is mappability removed? (#7)

Hi again, still no good news I am afraid. Today I tried to compute mappability for long reads after choosing 8kb as kmer length. This is apparently too much asking to the gem software since I was still at 0.3% of my data after 24h. Is there another method I could apply to compute a mappability track for a very repetitive plant genome using long kmers in the kb range (eg pacbio long-reads) One additional issue is very likely the platform error rate (CLR reads, not CCS) here which makes this even more difficult. Can someone comment on the idea and maybe suggest tools? Thanks

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHubhttps://eur01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fsmarco%2Fgem3-mapper%2Fissues%2F7%23issuecomment-638828523&data=01%7C01%7Cdan.smith%40rothamsted.ac.uk%7C217db431a4254d5389fa08d808865986%7Cb688362589414342b0e37b8cc8392f64%7C1&sdata=L0926e9tQBNpPzag2D2%2B1zPj67jctyDGZpOJs97xoM4%3D&reserved=0, or unsubscribehttps://eur01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fnotifications%2Funsubscribe-auth%2FALSETT7R4UE25TEAYHCSMLTRU6KONANCNFSM4E5XBP5A&data=01%7C01%7Cdan.smith%40rothamsted.ac.uk%7C217db431a4254d5389fa08d808865986%7Cb688362589414342b0e37b8cc8392f64%7C1&sdata=Hrf1AjC17fTzzxpto3axi8DD9NZb1s5FTn8vQvOEZF0%3D&reserved=0.

Rothamsted Research is a company limited by guarantee, registered in England at Harpenden, Hertfordshire, AL5 2JQ under the registration number 2393175 and a not for profit charity number 802038.

[splaisan]

thanks a lot @speromelior for the link, I remember I came across this tool earlier but somehow did not succeed to use it at the time

I was not meaning to compute on long reads but instead to compute reference mappability for read length in kb's, so using 'map -K 1000 -E 150' or something like that. Note sure genmap will digest such large values but I will try !

Thanks

[smarco]

Thanks, @speromelior for the feedback.

Old GEM-mappability was not built to support such read-lengths (i.e., 8Kbases), so we are in uncharted territory here: @splaisan, let us know your experience.

Cheers,