s-will/locarna

Alignment of RNAs

locarna's Issues

mlocarna realign of stk input

mlocarna realign seems to accept stockholm input with warnings but does not handle it properly

No colors in mLocARNA structure predictions from conda installation

Installation: conda install -c bioconda locarna
Version: LocARNA 1.9.1
Platform: Ubuntu 22.04 LTS

The RNA structures (alirna.ps) produced by mLocARNA do not contain any colors (example below). Any ideas/guidance on how to implement colors in the predicted structures would be appreciated.

Thank you!

mlocarna sparse option

Hi Sebastian,
There might be an error for mlocarna sparse option. The effect of adding -sparse option to mlocarna is currently only setting the pw aligner:
if ($sparse) { $pw_aligner = "$bindir/sparse"; }
However the in-loop probabilities will not be calculated by locarna-rnafold-pp by only setting the sparse option. If that is the case then opt_in_loop_probabilities should be automatically set if sparse option is given. Do you agree?
Cheers,

LocARNA uses the removed STL method

Build fails:

libtool: compile:  c++ -DHAVE_CONFIG_H -I. -I/usr/local/include -I/usr/local/include/ViennaRNA -pthread -O2 -pipe -fstack-protector-strong -fno-strict-aliasing -I/usr/local/include/catch2 -MT LocARNA/sequence.lo -MD -MP -MF LocARNA/.deps/sequence.Tpo -c LocARNA/sequence.cc  -fPIC -DPIC -o LocARNA/.libs/sequence.o
LocARNA/scoring.cc:57:29: error: no member named 'bind2nd' in namespace 'std'
                       std::bind2nd(std::minus<score_t>(), x));
                       ~~~~~^
LocARNA/scoring.cc:61:26: error: no member named 'bind2nd' in namespace 'std'
        m.transform(std::bind2nd(std::minus<score_t>(), x));
                    ~~~~~^
2 errors generated.

According to the docs std::bind2nd was removed in C++17.

Compilers that default to C++17 or C++20 fail to compile such code.

Version: 2.0.0
clang-16
FreeBSD 13.2

no structure annotation in stockholm file with 2.0.0RC8

Hi there,

when using mlocarna --stockholm <INPUT> with version 2.0.0RC8, the final stockholm alignment does not have the #=GC SS_cons annotation. I did the same command with version 1.9.2.1 and 1.9.2.3, and in both cases the annotation line was present.

For 2.0.0RC8, I have to explicitly call mlocarna --stockholm --consensus-structure alifold <INPUT> in order to get the same behavior. I was not able to see from the release note / manual, whether this is intended, thus I created this issue.

Greetings from Jena!
Kevin

mlocarna stumbles over `--width`

When running mlocarna I sometimes get the following error with version 1.9.2.3

/home/berntm/miniconda3/envs/[email protected]/bin/locarna_p: unrecognized option '--width'

Cannot compute match probabilities by /home/berntm/miniconda3/envs/[email protected]/bin/locarna_p input/fruA input/fdhA  --min-bm-prob=0.0005 --min-am-prob=0.0005  --indel -350 --indel-opening -500 --ribofit true --struct-weight 200 --tau 50 --max-diff-am 30 --max-diff 60 --width 60 --min-prob 0.0005 --write-basematch-probs _tmp_.bmps --write-arcmatch-probs _tmp_.amps  >/dev/null: No such file or directory.

error: implicit instantiation of undefined template

clang8 on FreeBSD 12 complains:

In file included from LocARNA/alphabet.hh:87:
LocARNA/alphabet.icc:27:22: error: implicit instantiation of undefined template 'std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> >'
        std::copy_n(s.begin(), N, this->begin());
                     ^
/usr/include/c++/v1/iosfwd:210:32: note: template is declared here
    class _LIBCPP_TEMPLATE_VIS basic_string;
                               ^

Cannot compute match probabilities

Hello!
I have been having an issue with mlocarna. :

/home/vanda/locarna-1.9.2.1/bin/mlocarna --probabilistic --iterations=10 --consistency-transformation --threads=20 --noLP cluster_10.fasta
terminate called after throwing an instance of 'LocARNA::failure'
  what():  ERROR: am prob (0:530,534) (0:553,560) -nan
Aborted (core dumped)
Cannot compute match probabilities by /home/vanda/locarna-1.9.2.1/bin/locarna_p input/LINF_340013500_L input/LINF_340013600_L  --min-bm-prob=0.0005 --min-am-prob=0.0005  --struct-weight 200 --max-diff-am 30 --min-prob 0.0005 --write-basematch-probs _tmp_30.bmps --write-arcmatch-probs _tmp_30.amps  >/dev/null: Perl exited with active threads:
        7 running and unjoined
        12 finished and unjoined
        0 running and detached

Not sure why i'm having this issue as the sequences do not contain gaps and are below 1000 nucleotides long. Any help would be greatly appreciate :)
Thank you!!!
Vanda
cluster_10.zip

Passing --LP gives the error: only one of the options --noLP and --LP can be defined at a time.

FASTA input:
I have tried numerous FASTAs (with 2 till 1000 sequences) an example is:

>test1
AAATCGATCAGCAGCATCAGCTAGCATCAGC
>test2
AGCGTAGCATCGATCGATCGACTAGCTACG

LocARNA version
LocARNA 1.9.2.1

Error
When I simply call: mlocarna test.fasta --LP, it always throws the error: Only one of the options --noLP and --LP can be defined at a time. (Notably, passing --noLP works fine)

Tests fail to compile: error: static assertion failed due to requirement '__libcpp_random_is_valid_urng<Catch::RandomNumberGenerator, void>::value'

In file included from test_locarna_lib.cc:2:
In file included from ./catch.hpp:74:
In file included from /usr/include/c++/v1/sstream:186:
In file included from /usr/include/c++/v1/istream:165:
In file included from /usr/include/c++/v1/ostream:168:
In file included from /usr/include/c++/v1/bitset:128:
In file included from /usr/include/c++/v1/string:551:
In file included from /usr/include/c++/v1/string_view:222:
In file included from /usr/include/c++/v1/algorithm:1851:
In file included from /usr/include/c++/v1/__algorithm/ranges_sample.h:13:
In file included from /usr/include/c++/v1/__algorithm/sample.h:18:
/usr/include/c++/v1/__random/uniform_int_distribution.h:235:5: error: static assertion failed due to requirement '__libcpp_random_is_valid_urng<Catch::RandomNumberGenerator, void>::value': 
    static_assert(__libcpp_random_is_valid_urng<_URNG>::value, "");
    ^             ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
/usr/include/c++/v1/__algorithm/shuffle.h:154:35: note: in instantiation of function template specialization 'std::uniform_int_distribution<long>::operator()<Catch::RandomNumberGenerator>' requested here
            difference_type __i = __uid(__g, _Pp(0, __d));
                                  ^
/usr/include/c++/v1/__algorithm/shuffle.h:166:14: note: in instantiation of function template specialization 'std::__shuffle<std::_ClassicAlgPolicy, std::__wrap_iter<Catch::TestCase *>, std::__wrap_iter<Catch::TestCase *>, Catch::RandomNumberGenerator &>' requested here
  (void)std::__shuffle<_ClassicAlgPolicy>(
             ^
./catch.hpp:6470:18: note: in instantiation of function template specialization 'std::shuffle<std::__wrap_iter<Catch::TestCase *>, Catch::RandomNumberGenerator &>' requested here
            std::shuffle( vector.begin(), vector.end(), rng );
                 ^
./catch.hpp:6488:44: note: in instantiation of function template specialization 'Catch::RandomNumberGenerator::shuffle<std::vector<Catch::TestCase>>' requested here
                    RandomNumberGenerator::shuffle( sorted );
                                           ^
1 error generated.

Version: 2.0.0
clang-15
FreeBSD 13.2

Get partition function of input sequences

Hi Sebastian,
For an structure scoring test I need the partition function (or even better the ensemble energy) of input sequences A & B. I was wondering if you already have extracted this information into some variable. The closest thing I could find in the code is min_free_energy_ entity of the class RnaEnsembleImpl.
If it's not already there could you guide how would be best to get and store that value?
Cheers

free_end_gaps parameter description ambiguous

Hello Sebastian,
i just wanted to let you know that I had to look in the source code to understand the format of the free_end_gaps parameter. I thought the "-" minus signs are just placeholder and tried to do things like: "R1L1" or "TTFF" to specify which end gaps I want to be free. The help could use a small update on this.

running local - two files.

Dear,

Sorry for my silly question. The software is already installed at the server.. it seems everything is ok. I decided to installed it because I have a thousand sequences to run..Instead, use the webserver.

Calling help from terminal "... , which
specify the two input sequences or input alignments"

Could you please give some tips to fix it?
Thanks in advance.

control size of output directory

Consider to write/keep large additional information only on demand
e.g. probs subdirectory containing all the single probabilities in mlocarna --probabilistic

Can this be changed (easily) such that writing to disk can be avoided?

provide distances in tree

Consider to (optionally?) provide distances in the guide tree

Pass energy parameter file to locarna_rnafold_pp

Hi @s-will,

What do you think of having optional parameter to pass an energy param file to locarna_rnafold_pp?
--paramFile=paramfile
I'll start implementing that, first wanted to know your opinion and be sure it is not already integrated inside locarna :)

Best,
M

please tag for releases

could you add tags for stable releases?

locarna --probabilistic fails to read arc scores

Hi dear team,

I am encountering an issue running locarnap on my fasta alignment: the program fails on reading arc scores from its generated files. Upon execution of the same command, the program will fail in the same step every time, but on different lines.

Additional details:
I am using version 1.9.2.3, installed with conda via bioconda.
I am running the program on on an Intel(R) Xeon(R) CPU E5-2630 v4 @ 2.20GHz machine
distribution: CentOS Linux release 7.3.1611

input:
Please find attached the input file I used, as well as the arc scores files which the program fails to read.

commands:
cd <input_path_dir>
mlocarna <input_path_dir> --probabilistic -tgtdir .

error:
terminate called after throwing an instance of 'LocAR
reproducing_data.zip
NA::failure'
what(): Cannot read arc match scores. Invalid line 9: 67 106 83 0.256865
sh: line 1: 624 Aborted (core dumped) */locarna input/JQ654586 input/EU627010 --struct-weight 200 --max-diff-am 30 --noLP --mea-alignment --min-prob 0.0005 --pp=intermediates/intermediate1.pp --clustal=intermediates/intermediate1.aln --read-match-probs=tmp.bmps --read-arcmatch-probs=tmp.amps > /dev/null
Cannot execute */locarna: Inappropriate ioctl for device

Could you please let me know how to fix it?

Thanks!

2.0.0-RC8 fails: No rule to make target 'exparna_p.1'

gmake[2]: Entering directory '/usr/ports/biology/locarna/work/LocARNA-2.0.0RC8/Doc'
Making all in man
gmake[3]: Entering directory '/usr/ports/biology/locarna/work/LocARNA-2.0.0RC8/Doc/man'
gmake[3]: *** No rule to make target 'exparna_p.1', needed by 'all-am'.  Stop.
gmake[3]: Leaving directory '/usr/ports/biology/locarna/work/LocARNA-2.0.0RC8/Doc/man'
gmake[2]: *** [Makefile:399: all-recursive] Error 1
gmake[2]: Leaving directory '/usr/ports/biology/locarna/work/LocARNA-2.0.0RC8/Doc'
gmake[1]: *** [Makefile:667: all-recursive] Error 1

FreeBSD 12 amd64

handling of empty input sequences

empty input sequences cause locarna to crash (violated assertion)

'make check' fails

hi!First of all, thanks for your software to help me to analysis my data.
Unfortunately，I meet some problem when i install LocaRNA, and my commends as follow:

wet https://github.com/s-will/LocARNA/releases/download/v1.9.2.1/locarna-1.9.2.1.tar.gz 
tar zxvf locarna-1.9.2.1.tar.gz 
cd locarna-1.9.2.1
./configure --prefix=my/located/locarna-1.9.2.1
make
make check

Before this, I have installed ViennaRNA-2.4.18， and my error is：
{{{
Testsuite summary for LocARNA 1.9.2.1
TOTAL: 2
PASS: 1
SKIP: 0
XFAIL: 0
FAIL: 1
XPASS: 0
ERROR: 0
See src/Tests/test-suite.log
Please report to [email protected]
make[6]: *** [Makefile:746: test-suite.log] Error 1
make[6]: Leaving directory '/home/xxf/lini/biosoft/LocARNA/locarna-1.9.2.1/src/T ests'
make[5]: *** [Makefile:854: check-TESTS] Error 2
make[5]: Leaving directory '/home/xxf/lini/biosoft/LocARNA/locarna-1.9.2.1/src/T ests'
make[4]: *** [Makefile:932: check-am] Error 2
make[4]: Leaving directory '/home/xxf/lini/biosoft/LocARNA/locarna-1.9.2.1/src/T ests'
make[3]: *** [Makefile:934: check] Error 2
make[3]: Leaving directory '/home/xxf/lini/biosoft/LocARNA/locarna-1.9.2.1/src/T ests'
make[2]: *** [Makefile:963: check-recursive] Error 1
make[2]: Leaving directory '/home/xxf/lini/biosoft/LocARNA/locarna-1.9.2.1/src'
make[1]: *** [Makefile:1112: check] Error 2
make[1]: Leaving directory '/home/xxf/lini/biosoft/LocARNA/locarna-1.9.2.1/src'
make: *** [Makefile:650: check-recursive] Error 1
}}}
i will give the src/Tests/test-suite.log in the website:
https://github.com/zxy1555847/-/projects/1
Thank you!

--noLP and --free-endgaps option results in segfault in specific scenario

Hello,

There appears to be some situlation specific bug in locarna 1.9.1.

Originally, I've come across it when I was running mlocarna with --free-endgaps parameter for 5 anchored sequences, which resulted in segfault. The situation did not happen when --free-endgaps were not set in mlocarna because the constructed guide tree was different.

After some digging, I've found that in actually the --noLP flag for locarna aligner together with --free-endgaps are responsible for segfault, when the mlocarna runs the progressive alignment. (The --noLP is set automatically by mlocarna for some reason.)

To recreate the bug I'm attaching a mlocarna output dir. The command in question as reported by --moreverbose is:
/usr/local/bin/locarna intermediates/intermediate1.pp intermediates/intermediate2.pp --free-endgaps="++++" --struct-weight 200 --max-diff-am 30 --noLP --min-prob 0.0005 --clustal=intermediates/intermediate3.aln --pp=intermediates/intermediate3.pp > /dev/null

The --noLP together with --free-endgaps result in segfault, others may be set to default.
So the minimal code is locarna --noLP --free-endgaps ++++ intermediate1.pp intermediate2.pp from the intermediates dir.

I've traced that the segfault is raised from aligner file, from:
AlignerImpl::trace_arcmatch_noLP method, more specifically:
alignment_.add_basepairA(arcAI.left(),arcAI.right());

line (1111 in master, 1028 in locarna-1.9.1).

I've tried to build Locarna from git, but did not succeed, so I can't provide any inside for more recent version of Locarna then 1.9.1.

file1.out.zip
I hope it helps with tracing the bug.

Best regards
Marek Schwarz

LocARNA output cuts the sequence position if the header is to long

Hello Sebastian,

today I noticed that if the sequence position of the header are 4 digits the displayed header in the output is missing the last two digits of the header.

One example call is:
/some/dir/locarna-1.9.1/bin/mlocarna --tgtdir /some/dir/ --max-diff-am '30' --indel -350 --indel-opening -500 --struct-weight 190 --tau 0 -min-prob '0.005' --write-structure --pw-aligner-options "--pos-output" ../BRALIBASE/k2/U6/U6.apsi-80.sci-86.no-1.raw.fa -moreverbose -v

which produced this line in the output:
Perform progressive alignment ...
Align input/U58510_1_7471_75 + input/Z17282.1_96-198 --> intermediates/intermediate1

However the BRALIBASE fasta file looks like this:

BRALIBASEk2/k2/U6/U6.apsi-80.sci-86.no-1.raw.fa:

U58510.1_7471-7569
gcguucacgcaucuuuaaaauuggaacgauacagagaagauuagcauggccccugcgcaaggaugacacgcauaauccgagauuagguuccuuuauuua
Z17282.1_96-198
gucccuucggggacauccgauaaaauuggaacgauacagagaagauuagcauggccccugcgcaaggaugacacgcacaaaucgagaaaugguccaaauuuuu

Thanks for your support and greetings,
Teresa

Anchor line dropped when --keep-sequence-order flag

Hi,
I've noticed, that mlocarna drops the #A1 anchor sequence (which is present in result.aln when anchors are given) when --keep-sequence-order flag is given. The #A2 line is retained but moved from bottom of the alignment to the top.

The stockholm output file seems unaffected, but it is not sorted.

Moreover, when --keep-sequence-order flag is given, the result_prog.aln contains all sequences and annotations, but it is not sorted, and the result.aln is sorted but in my case, one sequence is missing.

Please see attached files:
mlocarna.tar.gz

The issue with A1 anchor is repeatable on the Locarna web interface with snoRNAs predefined dataset. However, I didn't check if the whole sequence disappears as well in the result.aln.

Best regards
Marek Schwarz

Calling LocARNA global alignment unsing the --penalize parameter

Hello Sebastian,
we tried to call LocARNA using the penalize parameter, which automatic turns on the local mode. We also tried to enable the local mode explicitly (--sequ-local=False), which was rising an error message. Is there a way to use the penalty for global alignment?
Best,
Teresa

Undefined references to ViennaRNA symbols

I'm trying to package LocARNA for Guix but it seems that the libraries are not linked properly. I'm getting errors like this:

./.libs/libLocARNA-1.5.a(rna_ensemble.o): In function `LocARNA::RnaEnsembleImpl::ptype_of_admissible_basepair(unsigned long, unsigned long) const':
/tmp/guix-build-locarna-1.8.10.drv-0/LocARNA-1.8.10/src/LocARNA/rna_ensemble.cc:592: undefined reference to `no_closingGU'
./.libs/libLocARNA-1.5.a(rna_ensemble.o): In function `LocARNA::RnaEnsembleImpl::unpaired_in_loop_prob_ali(unsigned long, unsigned long, unsigned long) const':
/tmp/guix-build-locarna-1.8.10.drv-0/LocARNA-1.8.10/src/LocARNA/rna_ensemble.cc:745: undefined reference to `no_closingGU'
/tmp/guix-build-locarna-1.8.10.drv-0/LocARNA-1.8.10/src/LocARNA/rna_ensemble.cc:767: undefined reference to `no_closingGU'
./.libs/libLocARNA-1.5.a(rna_ensemble.o): In function `LocARNA::RnaEnsembleImpl::unpaired_in_loop_prob_noali(unsigned long, unsigned long, unsigned long) const':
/gnu/store/39yi0d07qcfhv312lv7bkc60wl5pzpyb-viennarna-2.2.8/include/ViennaRNA/interior_loops.h:361: undefined reference to `no_closingGU'
/gnu/store/39yi0d07qcfhv312lv7bkc60wl5pzpyb-viennarna-2.2.8/include/ViennaRNA/interior_loops.h:361: undefined reference to `no_closingGU'
./.libs/libLocARNA-1.5.a(rna_ensemble.o):/gnu/store/39yi0d07qcfhv312lv7bkc60wl5pzpyb-viennarna-2.2.8/include/ViennaRNA/interior_loops.h:359: more undefined references to `no_closingGU' follow
./.libs/libLocARNA-1.5.a(rna_ensemble.o): In function `LocARNA::RnaEnsemble::arc_external_prob(unsigned long, unsigned long) const':
/tmp/guix-build-locarna-1.8.10.drv-0/LocARNA-1.8.10/src/LocARNA/rna_ensemble.cc:1217: undefined reference to `exp_E_ExtLoop'
/tmp/guix-build-locarna-1.8.10.drv-0/LocARNA-1.8.10/src/LocARNA/rna_ensemble.cc:1231: undefined reference to `exp_E_ExtLoop'
./.libs/libLocARNA-1.5.a(rna_ensemble.o): In function `LocARNA::RnaEnsembleImpl::compute_Qm2_ali()':
/tmp/guix-build-locarna-1.8.10.drv-0/LocARNA-1.8.10/src/LocARNA/rna_ensemble.cc:518: undefined reference to `fold_constrained'
...

These symbols appear to be defined in ViennaRNA's libRNA. I've tried to build with ViennaRNA 2.2.7 and 2.2.7 and with a static build of LocARNA (--enable-link-static=yes) and without. I'm not familiar with libtool, so I don't know what's going on here or how to fix this.

I even tried to patch src/Makefile.am before bootstrapping, adding the full path to "libRNA.a" to the LDADD line, but without success.

Using gcc 5.4 with locarna version 2.0.0RC6

Installing locarna version 2.0.0RC6 I encountered an error by using g++ 5.4

muellert@muellert-ThinkPad-T450s:~/Software/locarna-2.0.0RC6$ make
Making all in src
make[1]: Verzeichnis „/home/muellert/Software/locarna-2.0.0RC6/src“ wird betreten
depbase=`echo LocARNA/alignment.lo | sed 's|[^/]*$|.deps/&|;s|\.lo$||'`;\
/bin/bash ../libtool  --tag=CXX   --mode=compile g++ -std=c++14 -DHAVE_CONFIG_H -I.   -pthread -I/usr/local/include -I/usr/local/include/ViennaRNA   -g -O2 -MT LocARNA/alignment.lo -MD -MP -MF $depbase.Tpo -c -o LocARNA/alignment.lo LocARNA/alignment.cc &&\
mv -f $depbase.Tpo $depbase.Plo
libtool: compile:  g++ -std=c++14 -DHAVE_CONFIG_H -I. -pthread -I/usr/local/include -I/usr/local/include/ViennaRNA -g -O2 -MT LocARNA/alignment.lo -MD -MP -MF LocARNA/.deps/alignment.Tpo -c LocARNA/alignment.cc  -fPIC -DPIC -o LocARNA/.libs/alignment.o
In file included from /usr/include/c++/5/random:38:0,
                 from /usr/include/c++/5/bits/stl_algo.h:66,
                 from /usr/include/c++/5/algorithm:62,
                 from LocARNA/infty_int.hh:8,
                 from LocARNA/scoring_fwd.hh:8,
                 from LocARNA/alignment.hh:12,
                 from LocARNA/alignment.cc:3:
/usr/include/c++/5/cmath: In function 'constexpr bool std::isnan(float)':
/usr/include/c++/5/cmath:627:27: error: call to non-constexpr function '__float128 std::__builtin_isnan(__float128)'
   { return __builtin_isnan(__x); }
                           ^
/usr/include/c++/5/cmath: In function 'constexpr bool std::isnan(double)':
/usr/include/c++/5/cmath:631:27: error: call to non-constexpr function '__float128 std::__builtin_isnan(__float128)'
   { return __builtin_isnan(__x); }
                           ^
/usr/include/c++/5/cmath: In function 'constexpr bool std::isnan(long double)':
/usr/include/c++/5/cmath:635:27: error: call to non-constexpr function '__float128 std::__builtin_isnan(__float128)'
   { return __builtin_isnan(__x); }
                           ^
Makefile:890: die Regel für Ziel „LocARNA/alignment.lo“ scheiterte
make[1]: *** [LocARNA/alignment.lo] Fehler 1
make[1]: Verzeichnis „/home/muellert/Software/locarna-2.0.0RC6/src“ wird verlassen
Makefile:657: die Regel für Ziel „all-recursive“ scheiterte
make: *** [all-recursive] Fehler 1
muellert@muellert-ThinkPad-T450s:~/Software/locarna-2.0.0RC6$ g++ 
g++: fatal error: no input files
compilation terminated.
muellert@muellert-ThinkPad-T450s:~/Software/locarna-2.0.0RC6$ g++ --version
g++ (Ubuntu 5.4.0-6ubuntu1~16.04.9) 5.4.0 20160609
Copyright (C) 2015 Free Software Foundation, Inc.
This is free software; see the source for copying conditions.  There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.

One test fails: ERROR: Programm t_coffee cannot be found in search path!

ERROR: Programm t_coffee cannot be found in search path!
    Please install and/or set path.

Version: 2.0.0
FreeBSD 13.2

installing to 'lib/perl' not working in bioconda ..

Hi Sebastian,
installing the MLocarna.pl to lib/perl via

LocARNA/lib/perl/Makefile.am

Line 3 in fe6d478

perllibdir=$(prefix)/lib/perl

makes it invisible in bioconda packaging, since conda currently only activates the following pathes in @INC:

$(prefix)/lib/perl5/site_perl/5.22.0/x86_64-linux-thread-multi 
$(prefix)/lib/perl5/site_perl/5.22.0 
$(prefix)/lib/perl5/5.22.0/x86_64-linux-thread-multi 
$(prefix)/lib/perl5/5.22.0

could you please either change the installation directory or the bioconda package recipe to get it installed into one of the @INC folders?

Thanks,
Martin

"Can not read alignment file results/result.aln" in refinement of alignment

I am using mlocarna 2.0.0RC8 (installed from bioconda) to refine an alignment with structural constraints, for which I also have a guide tree. The alignment contains 736 sequences with a (gapped) length of 3042. (This is 5.8S, ITS2, and part of 28S rRNA).

Command line:

mlocarna --treefile 32S_guide.tree --realign 32S_ASVs.aln --cpus 3 -v

Output:

mLocARNA --- multiple Local (and global) Alignment of RNA --- LocARNA 2.0.0RC8
Copyright Sebastian Will

Locarna parameter for progressive alignment: -v --struct-weight 200 --max-diff-am 30 --noLP --consensus-structure none --min-prob 0.001
Locarna parameter for guide tree construction: -v --struct-weight 200 --max-diff-am 30 --noLP --consensus-structure none --min-prob 0.001
Warning: 32S_ASVs.out exists already.
Compute using 3 threads.
Compute pair probs ...
Read guide tree from file. 
Perform progressive alignment ...
Can not read alignment file results/result.aln

I have confirmed that the pair probabilities files were written in the output directory, and the file results/result.aln is absent, although there is a results/result.tree. I assume the alignment should have been generated by some previous step, but this doesn't seem to have happened.

missing consensus structure in stockholm

[2.0.0RC7] mlocarna does not write consensus structure in final (and intermediate)
stockholm alignment files.
It seems that there is no way to turn this on in mlocarna!

The Hit line of locarna global alignment sometimes gives wrong positions

Hello Sebastian,
today I noticed that some LocARNA calls I did generated wrong positions in the HIT line of the output.

one example call I did is:
/some/dir/locarna-1.9.1/bin/mlocarna --tgtdir /some/dir/ --max-diff-am '30' --indel -350 --indel-opening -500 --struct-weight 220 --tau 0 -min-prob '0.005' --write-structure --pw-aligner-options "--pos-output" ../BRALIBASE/k2/U6/U6.apsi-82.sci-68.no-1.raw.fa -moreverbose -v

and it generated The Hit line:
HIT 9344 1 18446744073709551615 107 101
But the line should look like this:
HIT 9344 1 1 107 101

Best wishes,
Teresa

spurious output

mlocarna outputs intermediary alignments in proabilistic mode ---> remove

Tracking file Packaging/debian/control

The configure/make process modifies the file 'Packaging/debian/control'
This generates a conflict error complain whenever a pull is invoked.
Error message:
error: Your local changes to the following files would be overwritten by checkout: Packaging/debian/control Please, commit your changes or stash them before you can switch branches. Aborting

No rule to make target 'exparna_p.1', needed by 'all-am'. Stop.

Fails in the FreeBSD ports framework:

Making all in Doc
gmake[2]: Entering directory '/usr/ports/biology/locarna/work/LocARNA-2.0.0RC7/Doc'
Making all in man
gmake[3]: Entering directory '/usr/ports/biology/locarna/work/LocARNA-2.0.0RC7/Doc/man'
gmake[3]: *** No rule to make target 'exparna_p.1', needed by 'all-am'.  Stop.
gmake[3]: Leaving directory '/usr/ports/biology/locarna/work/LocARNA-2.0.0RC7/Doc/man'
gmake[2]: *** [Makefile:399: all-recursive] Error 1
gmake[2]: Leaving directory '/usr/ports/biology/locarna/work/LocARNA-2.0.0RC7/Doc'
gmake[1]: *** [Makefile:667: all-recursive] Error 1
gmake[1]: Leaving directory '/usr/ports/biology/locarna/work/LocARNA-2.0.0RC7'
===> Compilation failed unexpectedly.

empty names in input

thanks to Paul Klemm:

RNAalifold, locarna

SOURCE of error : fasta with an empty genename

-> RNAalifold throws malloc error

-> locarna ignores gene with empty name

##############################################

##########

RNAalifold

##########

(generated the example.aln with 'clustalw2 example.fasta -CLUSTERING=UPGMA')

$ RNAalifold example.aln

RNAalifold(96372,0x7fffb166b380) malloc: *** error for object 0x7ff24c501cb0: pointer being freed was not allocated
*** set a breakpoint in malloc_error_break to debug
Abort trap: 6

##########

mlocarna

##########

ignores the gene with empty name '>' since it cannot create a file with name ''

locarnate: Invalid CODE attribute

After successfully installation, the command locarna is OK, but the command locarnate fails:

[zhut@cotton locarna-1.9.2.1]$ locarnate
Invalid CODE attribute: prototype($) at /usr/local/bin/../lib/perl/MLocarna/Aux.pm line 89.
BEGIN failed--compilation aborted at /usr/local/bin/../lib/perl/MLocarna/Aux.pm line 89.
Compilation failed in require at /usr/local/bin/../lib/perl/MLocarna/MatchProbs.pm line    @@@19.
BEGIN failed--compilation aborted at /usr/local/bin/../lib/perl/MLocarna/MatchProbs.pm line 19.
Compilation failed in require at /usr/local/bin/../lib/perl/MLocarna.pm line 12.
BEGIN failed--compilation aborted at /usr/local/bin/../lib/perl/MLocarna.pm line 12.
Compilation failed in require at /usr/local/bin/locarnate line 96.
BEGIN failed--compilation aborted at /usr/local/bin/locarnate line 96. # #

alingment ID in STOCKHOLMs output

Hello,

I used mlocarna to align tRNA isoforms. All works fine as far as I can tell, but the STOCKHOLM format files do not contain identifiers (which is to be expected).

I am getting:

# STOCKHOLM 1.0
#=GF CC Generated by LocARNA 2.0.0
#=GF SQ 26

tRNA-Ala-AGC-6-1   GGGGGUAUAGCUCA<snip>

what I need is:

# STOCKHOLM 1.0
#=GF ID tRNA-Ala-AGC

tRNA-Ala-AGC-1-1    GGGGGUATAGCUCAGU<snip>

Would you consider adding an extra feature/command line switch, i.e.:

 mlocarna --stockholm --threads 6 --name tRNA-Ala-AGC  tRNA-Ala-AGC.fa

Many thanks for your help

Darek Kedra

Cannot find pkgconfig file RNAlib2.pc

While running ./configure --with-vrna=/path/to/ViennaRNA-2.2.10 I get :

[...]

configure: error: =======================================================
Cannot find pkgconfig file RNAlib2.pc for the Vienna RNA package (>= 2.3.1)
in '/path/to/ViennaRNA-2.2.10/lib/pkgconfig'.
LocARNA REQUIRES the Vienna RNA package (>= 2.3.1);
It is freely available at http://www.tbi.univie.ac.at/RNA/.

yet find /path/to/ViennaRNA-2.2.10 -name "RNAlib2.pc" gives

/path/to/ViennaRNA-2.2.10/lib/pkgconfig/RNAlib2.pc

I also tried with export PKG_CONFIG_PATH=/path/to/ViennaRNA-2.2.10/lib/pkgconfig/
but this didn't help. Ideas? I am using locarna-1.9.1.tar.gz from http://www.bioinf.uni-freiburg.de/Software/LocARNA/

'make check' fails at locarnate 'Calculate multiple alignment...'

Running make check with the recent 1.9.2.3 release fails with the following error when testing for proper function of locarnate:

============================================================
TEST locarnate
CALL locarnate ./../../Data/Examples/archaea.fa

Calculate pair probabilities... done!
Calculate pairwise alignment with locarna 21/21 (Seq 6 <-> Seq 7)... done!      
Parse pairwise alignments... done!
Calculate multiple alignment...
t_coffee  -in=Ltest_results/input.lib -outfile=test_results/mult/tcoffee.aln 1>test_results/mult/tcoffee.out 2>test_results/mult/tcoffee.err
Calculate multiple alignment... done!
diff: test_results/mult/tcoffee.aln: No such file or directory
==================== DIFFERENT

I have t-coffee installed and all other checks seem fine