Dear friends: As a novice, I want to use Fedora to do analysis of RNA-seq results by using TopHat and Cufflinks softwares. What kind of Fedora linux software should I download? Please give me some advice!
Hi Liu,
These 2 are not in the repo IIRC, tophat is already in my local repo but the latter one you mentioned is not. I recently started packaging NLP/RNA softwares, if possible I can notify you when I'm ready.
Thanks.
Hi,
regarding tophat I would like to mention that in Debian we had some trouble to build it with latest seqan library. This was solved recently and patches are forwarded upstream (and available in Debian packaging repository - feel free to ping me for more advise if you have no clue how to find this).
Cufflinks is packaged for Debian as well. If you would like to see a full overview about software useful in life sciences in Debian where you can sneak into the packaging for patches (links to the packaging repository are given in the metainformation of these pages) feel free to have a look here:
http://blends.debian.org/med/tasks/bio http://blends.debian.org/med/tasks/bio-dev
I'd welcome better contact between Fedora scitech and Debian Med (debian-med@lists.debian.org) and Debian Science (debian-science@lists.debian.org).
Kind regards
Andreas.
On Wed, Jul 30, 2014 at 09:14:41AM +0800, Christopher Meng wrote:
Hi Liu,
These 2 are not in the repo IIRC, tophat is already in my local repo but the latter one you mentioned is not. I recently started packaging NLP/RNA softwares, if possible I can notify you when I'm ready.
Thanks.
scitech mailing list scitech@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/scitech
On Jul 30, 2014 2:24 PM, "Andreas Tille" andreas@an3as.eu wrote:
Hi,
First I don't use Debian for years, so I don't know how you deal with it. But anyway I think your effect is helpful at least.
Second, I didn't submit it to the review queue because in the past seqan was not available in Fedora until recent review from Antonio. Thus I bundled seqan in the package. If I push it to Fedora, I need to unbundle it and in this morning I failed because of the too new seqan.
And bowtie2 is still not packaged as well, although it's not a must have dep so far.
I'd like wait for a while until upstream has fixed it.
Thanks.
Hi,
On Wed, Jul 30, 2014 at 04:58:32PM +0800, Christopher Meng wrote:
First I don't use Debian for years,
That's perfectly OK.
so I don't know how you deal with it.
If you look at the links I've posted, in on the right side of the header of each package you will find a link to the packaging repository. Inside the debian/patches directory you will find potentially helpful patches to use also for Fedora.
But anyway I think your effect is helpful at least.
I hope so.
Second, I didn't submit it to the review queue because in the past seqan was not available in Fedora until recent review from Antonio. Thus I bundled seqan in the package. If I push it to Fedora, I need to unbundle it and in this morning I failed because of the too new seqan.
Hint for seqan: The build of the SeQan tools needs a real lot of memory. It could be a solution to provide only the header files in a package which might be sufficient to build tophat.
And bowtie2 is still not packaged as well, although it's not a must have dep so far.
As you might have noticed bowtie2 is also packaged for Debian. I have no idea whether different architectures in Fedora are an issue but the Debian patch for this package contains better detection of the architecture.
Kind regards
Andreas.
Andreas Tille andreas@an3as.eu writes:
Hint for seqan: The build of the SeQan tools needs a real lot of memory. It could be a solution to provide only the header files in a package which might be sufficient to build tophat.
I don't think there's a need since it builds in the Fedora infrastructure, or maybe not copr, but I think it did before. As far as I remember, it was just the test programs that were the problem.
And bowtie2 is still not packaged as well, although it's not a must have dep so far.
As you might have noticed bowtie2 is also packaged for Debian. I have no idea whether different architectures in Fedora are an issue but the Debian patch for this package contains better detection of the architecture.
I'll look, thanks. I don't remember what that was about.
Christopher Meng cickumqt@gmail.com writes:
On Jul 30, 2014 2:24 PM, "Andreas Tille" andreas@an3as.eu wrote:
Hi,
First I don't use Debian for years, so I don't know how you deal with it. But anyway I think your effect is helpful at least.
You'd find tophat as https://tracker.debian.org/pkg/tophat. Search at http://packages.debian.org/ if you don't know the exact name.
Second, I didn't submit it to the review queue because in the past seqan was not available in Fedora until recent review from Antonio. Thus I bundled seqan in the package. If I push it to Fedora, I need to unbundle it and in this morning I failed because of the too new seqan.
I pointed that out on the review. We definitely need to avoid duplication. I seem to remember there's a wiki page where work in progress could be recorded.
And bowtie2 is still not packaged as well, although it's not a must have dep so far.
I should prod Adam if he's still interested enough to do the work and isn't listening https://bugzilla.redhat.com/show_bug.cgi?id=790629
I'd like wait for a while until upstream has fixed it.
I built it here against my version of seqan-1.4, but that failed to build in copr, which I can't deal with now. I haven't ported the official seqan to EPEL6 yet.
I'll revisit this stuff in 10 days or more.
Andreas Tille andreas@an3as.eu writes:
Hi,
regarding tophat I would like to mention that in Debian we had some trouble to build it with latest seqan library. This was solved recently and patches are forwarded upstream (and available in Debian packaging repository - feel free to ping me for more advise if you have no clue how to find this).
That was the problem with the recent fedora seqan packaging. I used the bundled version, but I'll check the Debian changes, which weren't there when I looked originally. Thanks for the info.
I'd welcome better contact between Fedora scitech and Debian Med (debian-med@lists.debian.org) and Debian Science (debian-science@lists.debian.org).
For what it's worth, if I have something non-trivial to package -- which I try to do with everything free-ish that's installed on our cluster -- I look for Debian packaging to start with (and possibly OpenSuSE science, but a lot of that isn't too clean). I'm typing this at a Debian desktop...
Christopher Meng cickumqt@gmail.com writes:
Hi Liu,
These 2 are not in the repo IIRC, tophat is already in my local repo but the latter one you mentioned is not. I recently started packaging NLP/RNA softwares, if possible I can notify you when I'm ready.
I should find time to talk about that again when I get back from holiday. Not everything we're using is currently published or yet in a fit state for Fedora, but I'm gradually working on it.
刘定干 dgliu@sibcb.ac.cn writes:
Dear friends: As a novice, I want to use Fedora to do analysis of RNA-seq results by using TopHat and Cufflinks softwares.
cufflinks appears to be non-free, in which case it can't go into Fedora, though there doesn't seem to have been a ruling from fedora-legal on it.
What kind of Fedora linux software should I download? Please give me some advice!
Sorry, I don't have time to explain what to do with this, but maybe someone else can:
You should be able to rebuild the tophat rpm under http://copr.fedoraproject.org/coprs/loveshack/livhpc/ on anything RHEL6-compatible or above, but on EL6 it needs an updated boost, e.g. from that repo. I'll try to start a fedora20 build anyhow.
A cufflinks rpm can be built from http://arc.liv.ac.uk/downloads/misc/SPECS/cufflinks.spec. The current bowtie is in the same places.
Hi,
On Wed, Jul 30, 2014 at 11:42:03AM +0100, Dave Love wrote:
RNA-seq results by using TopHat and Cufflinks softwares.
cufflinks appears to be non-free, in which case it can't go into Fedora, though there doesn't seem to have been a ruling from fedora-legal on it.
The "non-free-ness" in Debian is caused by the use of locfit. I'm quoting the relevant section from debian/copyright:
Files: src/locfit/* Copyright: © 1996-2001 Lucent Technologies, Bell Laboratories. © 1996-2001 Catherine Loader. License: Locfit ABOUT ALGORITHMS: The Locfit implementation has been designed largely from the point of view of generality: to provide a set of functions that can be used for as wide a range of local fitting problems as possible. It is not intended to be the fastest or most efficient implementation possible. In addition to generality, the code in many places makes extensive trade-offs made between speed and numerical accuracy (some of which can be controlled through optional arguments). Many of Locfit's options will only be used in a small fraction of cases; for other cases, they add to the cost of overhead (an obvious example is multi-dimensional fitting: when used in 1-d, many loops reduce to for(i=0;i<1;i++)). Additionally, the user interfaces (i.e. the R and S-Plus code) add significantly to the computational overhead. For these reasons, the Locfit code, as distributed, should not, and can not, be used to derive meaningful benchmarks, either for the speed or accuracy of algorithms. Anyone wishing to use Locfit in any kind of comparative benchmark study must contact and obtain permission from the Author. . COPYRIGHT: . Copyright (c) 1996-2001 Lucent Technologies, Bell Laboratories. SCB code is Copyright (c) 1996-2001 Jiayang Sun. . Permission to use, copy, modify, and distribute this software for any purpose (with the exceptions noted in `About Algorithms' above) without fee is hereby granted, and provided that this entire notice is included in all copies of any software which is or includes a copy or modification of this software and in all copies of the supporting documentation for such software. . THIS SOFTWARE IS BEING PROVIDED "AS IS", WITHOUT ANY EXPRESS OR IMPLIED WARRANTY. IN PARTICULAR, NEITHER THE AUTHOR NOR LUCENT TECHNOLOGIES MAKE ANY REPRESENTATION OR WARRANTY OF ANY KIND CONCERNING THE MERCHANTABILITY OF THIS SOFTWARE OR ITS FITNESS FOR ANY PARTICULAR PURPOSE. Comment: the files contain “See README file for details.” and the license displayed above is from the README file of the locfit source available at http://stat.bell-labs.com/project/locfit/dist/locfit.tgz Source: http://stat.bell-labs.com/project/locfit
So while the code is free software the restriction:
... should not, and can not, be used to derive meaningful benchmarks, either for the speed or accuracy of algorithms.
is something what is regarded as non-free by Debian. I several times tried to contact the authors about this but never got any response.
I'd be interested about the opinion of fedora-legal.
Kind regards
Andreas.
Andreas Tille andreas@an3as.eu writes:
Hi,
On Wed, Jul 30, 2014 at 11:42:03AM +0100, Dave Love wrote:
RNA-seq results by using TopHat and Cufflinks softwares.
cufflinks appears to be non-free, in which case it can't go into Fedora, though there doesn't seem to have been a ruling from fedora-legal on it.
The "non-free-ness" in Debian is caused by the use of locfit.
Yes https://bugzilla.redhat.com/show_bug.cgi?id=694998
So while the code is free software the restriction:
... should not, and can not, be used to derive meaningful benchmarks, either for the speed or accuracy of algorithms.
is something what is regarded as non-free by Debian.
It's a restriction usage, if taken at face value, contra Freedom 0, in which case it's not free software.
I several times tried to contact the authors about this but never got any response.
I couldn't track down whoever seemed to be the relevant one, but I didn't try too hard. It's useful to know it's not worth the effort. I did get a response from the cufflinks maintainer, but I think the missing licence file didn't get added to the cufflinks distribution.
I'd be interested about the opinion of fedora-legal.
I can ask after I get back from holiday, unless someone else cares to. I don't know if you need to be a Fedora developer to do that, though.
I had hoped it might be possible to set up a repo that could contain the non-free things we have to install (abyss is another in this area). Unfortunately the UK HPC community isn't interested in that sort of thing -- or at least the people that matter -- and I don't have the resources here.
scitech@lists.fedoraproject.org