[Bioperl-l] error during installation
Fields, Christopher J
cjfields at illinois.edu
Sun Jun 17 01:21:36 EDT 2012
On Jun 16, 2012, at 5:15 PM, Peng Du wrote:
> Thank you for all your replies.
> So,what do you suggest me do if I want to do alignment with external tools like clustawl or dpalign? In the current distribution of BioPerl, ext package is still required to do this, I am afraid.
No, you should only need bioperl-run to run clustalw, and bioperl core modules to parse it. If you have anything from those packages that is failing from a lack of bioperl-ext we need to know about it (preferably with code examples and errors encountered).
> Thank you.
> Sent from my iPhone
> On 2012/06/17, at 2:43, "Fields, Christopher J" <cjfields at illinois.edu> wrote:
>> On Jun 16, 2012, at 3:50 AM, Leon Timmermans wrote:
>>> On Sat, Jun 16, 2012 at 5:28 AM, Du, Peng <du at ibio.jp> wrote:
>>>> Hi everybody,
>>>> I am a newbie, I encountered a problem while installing bioperl-ext.
>>>> The error message is :
>>>> /usr/bin/ld: libs/libsw.a(aln.o): relocation R_X86_64_32 against
>>>> `.rodata.str1.8' can not be used when making a shared object;
>>>> recompile with -fPIC
>>>> libs/libsw.a: could not read symbols: Bad value
>>>> collect2: ld returned 1 exit status
>>>> I googled some solutions, but all of them are confusing, could you
>>>> tell me where exactly should I add the -fPIC? In the makefile under
>>>> the bioperl-ext-1.4 directory? I tried to add this to the CCCDLFLAGS,
>>>> CCDLFLAGS, LDDLFLAGS and LDFLAGS in that file. But it still doesn't
>>>> Could you tell me what to do or where could I find the answer?
>>>> Thank you very much.
>>>> Best Regards,
>>> You should not install it in the first place. BioPerl-ext has been
>>> partly merged into the main Bioperl distribution, and partly spun off
>>> into its own dists (Bio::Tools::HMM, not released yet AFAIK). You
>>> should just install the former, or help release the latter.
>> For all intents and purposes I would consider all parts (past and present) of bioperl-ext deprecated, unless we can find someone willing to take them on. And, with that in mind, frankly I would be much more amenable to supporting XS (or similar) interfaces to common bioinformatics libraries (Bio::DB::Sam as an example) than ones specifically designed with bioperl in mind. The latter (bioperl-ext) have significantly bit-rotted to the point I'm not sure whether they would be worth keeping around.
More information about the Bioperl-l