File hmmer.changes of Package hmmer3
-------------------------------------------------------------------
Thu Nov 17 06:16:16 UTC 2022 - Stefan Brüns <stefan.bruens@rwth-aachen.de>
- Update to version 3.3.2
* Fixed a recently introduced bug that could cause hmmsearch
(and presumably hmmscan) to segfault on rare comparisons
involving highly biased sequences.
* ./configure –enable-PIC wasn’t setting -fPIC option in
impl-sse, impl-vmx Makefiles.
* fixed an uninitialized ptr in makehmmerdb, in the fm_data
structure, which could cause makehmmerdb to crash.
-------------------------------------------------------------------
Thu Nov 21 12:11:17 UTC 2019 - Luigi Baldoni <aloisio@gmx.com>
- Update to version 3.3
most important changes:
* We improved the `hmmpgmd` search daemon. (`hmmpgmd` is the
compute server used at EBI to support their HMMER web
servers.) Now `hmmpgmd` handles large target sequence
databases more efficiently by using "sharding". Instead of
loading the entire target sequence database into memory on
every node, a target database can be sharded across multiple
nodes. The `hmmpgmd` daemon also now has its own user guide.
* We improved how we calculate our default sequence weights
(Henikoff position-based weights), especially on deep
alignments of 10-100K+ sequences. Now we calculate PB weights
only on consensus columns, not all columns. This avoids some
cases of artifactually extreme weights on very gappy
alignments. We also changed to a better rule for defining
sequence fragments. These changes give a small improvement
in sensitivity/specificity benchmarking of HMMER/Pfam
searches, and substantial speed improvements in building
profiles on deep alignments. Because of these changes,
profile HMMs built with version 3.3 give slightly different
scores compared to previous HMMER3 versions.
* Fixed a bug in the `hmmstat` "compKL" (composition KL
divergence) calculation, which was off by a constant. Now it
is reported as a standard KL divergence in bits.
bug fixes:
* fixed a bug where in some rare and complicated situations,
`nhmmer` could report overlapping envelopes on reverse
strand. [iss#159]
* Several bugs were fixed in MPI mode, including iss#157
(`hmmsim --mpi` was always segfaulting, even on simple
examples) and iss#154 (problems in `hmmscan` that were
corrupting output data).
* Fixed some 32-bit integer overflow bugs in `hmmpgmd`.
* Fixed a bug in the `hmmstat` "compKL" (composition KL
divergence) calculation, which was off by a constant. Now it
is reported as a standard KL divergence in bits.
* Fixed a bug where `hmmconvert --outfmt` wasn't recognizing
`3/f` as a format.
Smaller changes:
* Our fasta format parser now detects aligned FASTA format
(.afa files) more robustly, and will not attempt to read a
.afa file as an unaligned sequence file. [iss#153]
* Our `make check` tests depend on Python >= 3.5. Added checks
in `./configure` and `make` to fail gracefully if python3
isn't available.
* `./configure` now always calls `AC_PROG_CC_STDC` to add
compiler flags for C99 (even with icc).
* Removed undocumented `--{tmm,tmi,tmd,tim,tii,tdm,tdd}`
options from hmmbuild.
* Data serialization routines (used in `hmmpgmd`) were
rewritten and improved.
- Spec cleanup
-------------------------------------------------------------------
Fri Jun 15 08:34:00 UTC 2018 - aloisio@gmx.com
- Update to version 3.2.1
* iss #142: compilation fails ungracefully on unsupported
ppc64le
* iss #140: jackhmmer --fast segfault. Removes --fast
and --symfrac options from jackhmmer and its documentation.
* iss #141, #120: compilation failures on i586, ppc64be.
* iss #132: hmmbuild miscounts transitions in final node.
-------------------------------------------------------------------
Fri Jun 8 17:05:33 UTC 2018 - aloisio@gmx.com
- Update to version 3.2
Larger changes:
* HMMER is now distributed under the BSD license, instead of
the GPL.
* The 'make install' has been streamlined. Now we install only
programs and man pages, for 18 HMMER programs and 22 Easel
miniapps. We don't install development .h headers or .a
libraries.
* We changed our default policy on the number of threads/cores
that the search programs use. Previously multithreaded
programs would use all available cores by default; now we use
two worker threads by default (~2 cores), if multiple cores
are available. HMMER3 search programs do not scale much beyond
2 cores anyway, due to input saturation and memory use.
* H3 now strictly requires either SSE2 or Altivec/VMX vector
support. The portable 'dummy' implementation has been removed.
This was an non-vectorized portable pure C implementation,
much slower than HMMER3 on x86 and PowerPC platforms with SSE2
or Altivec/VMX vector instructions. We had included it because
we could -- HMMER3 implements "generic" non-vectorized versions
of its core algorithms for unit testing purposes anyway. We
thought maybe it'd be useful. As far as I can tell, it was
never useful, but several packagers on non-x86, non-PowerPC
platforms packaged and distributed it, ignoring the prominent
health warnings we put on it.
Smaller new features include:
* improved documentation of --mxfile and score matrix file
format
* adds support for EBI's "uncheck all hits" on jackhmmer web
site
* removed experimental hmmbuild --seq_weights_r and
--seq_weights_e options
* appeased new gcc6 --Wmisleading-indentation warning
* SSI indexing is now robust against duplicate keys
* autoconf now robust against someone running gnu `autoheader`
* improved autoconf of SIMD vector support
* improved autoconf of DAZ,FTZ support on x86 processors
* improved autodetection of phylip interleaved vs sequential
format
* improved autodetection of A2M vs. aligned FASTA format
* clarified A2M format parse error when file contains 'O'
residue
* improved portability to Intel/Solaris
* FASTA parser now ignores description line after a ctrl-A
* MSAs output with -A option now include name, acc, desc,
author markup.
* In Easel miniapps, esl-translate replaces previous (buggy)
esl-stranslate.
* Easel miniapps include esl-alirev, reverse complementing
RNA/DNA alignment
* nhmmer/nhmmscan now allow alignments as target dbs
* target relentropy for DNA models changed from 0.45 to 0.62
* updated the User Guide
Fixed bugs include:
* nhmmscan faulted on old model files because it expected MAXL
field
* --cut-ga wasn't checking that GA line was actually present
in model
* GA thresholds weren't being captured from DNA/RNA Stockholm
alignments
* A2M alignment parsing was corrupted on lines ending in
insertions
* alphabet-guessing failed for FASTA seq >4096 bytes and <4000
residues
* genbank to fasta format conversion was leaving extra \n in
accession
* NCBI db format: multithreaded nhmmer was failing to read
* NCBI db format: 1st database sequence was skipped for second
query (#e6)
* unit test failures in esl_normal, esl_random due to floating
point math
* nhmmer was corrupted by * symbols in input sequence (iss#118)
* hmmsearch -A crashed if top-ranked hit has no domains
(iss#131)
* new variety of "backconverted subseq" hmmsearch crash on *'s
(iss#135)
* qsort() callbacks now return -1/0/1 as they should.
(Infernal iss#11)
* esl_buffer hangs when input ends with \r (Easel iss#23)
- Dropped hmmer-bufferoverflowstrncat.patch (merged upstream)
- Dropped devel subpackage (easel no longer installs header files
and static library)
-------------------------------------------------------------------
Wed Oct 19 08:13:27 UTC 2016 - aloisio@gmx.com
- Update to 3.1b2 (see release-notes for changelog)
- Spec cleanup
- Fixed ppc64/ppc64le build
- Backported hmmer3-bufferoverflowstrncat.patch from trunk
to avoid buffer overflow errors.
- Added devel subpackage
-------------------------------------------------------------------
Sun Mar 20 08:49:09 UTC 2016 - mailaender@opensuse.org
- Adapted for openSUSE science:
* spec file cleanup
* use source urls
* fix rpmlint warnings
* don't strip binaries manually
-------------------------------------------------------------------
Fri Sep 10 11:48:55 UTC 2010 - kai.blin@biotech.uni-tuebingen.de
- Initial spec file heavily based on Dominik Borowski's version
o Fixed a couple of rpmlint warnings
