[mira_talk] MIRA V3.2.1 available

From: Bastien Chevreux <bach@xxxxxxxxxxxx>
To: mira_announce@xxxxxxxxxxxxx, mira_talk@xxxxxxxxxxxxx
Date: Tue, 30 Nov 2010 22:28:54 +0100
Dear all,

MIRA V3.2.1 has been released at SourceForge (since Sunday already, but that 
is another topic which will have its own post on MIRA talk today).

The source and a package with 64 bit binaries can be found here:

  http://sourceforge.net/projects/mira-assembler/

In the past few weeks I've included most of the feedback I got for the 3.2.0 
and the 3.2.1 release candidate version (rc1 & 2) plus a couple of 
improvements which I had on my list since quite some time. Most notable 
changes:

Improvements:
- slowdowns in the SW alignment phase while running over NFS has been
  eliminated. This single most performance breaking part with slowdowns >>15x
  which are now away.
- considerable speed-ups in different phases of an assembly when working with
  millions of reads (>5m) both in mapping or in de-novo mode, for EST and
  genome assemblies.
- SMALT (successor of SSAHA2) result files can now also be parsed to clip
  reads with -CL:msvs.
- simplified defining strain names when working with only one name, now no
  need to create strain data files in this case.
- as users often forgot to do that: default "--job" parameters for mapping now
  include default strain names for reference sequences and mapped reads so
  that subsequent SNP analysis or sequence extraction routines can work
  optimally.
- reduction of maximum memory footprint
- improved default parameters for some mapping assemblies (with Solexa data)
- documentation reworked in parts to simplify walk-throughs (e.g. Solexa)

Bugfixes
- miraSearchESTSNPs step 2 & 3 now work again as advertised
- hopefully fixed multi-threading problem in SKIM, leading to subsequent
  aborts of MIRA when errors were detected.
- fixed possible out-of-memory error for very small assemblies or assemblies
  where all reads were clipped away by MIRA.
- lots of other small things

Feedback is - as always - much appreciated.

Best,
  Bastien

-- 
You have received this mail because you are subscribed to the mira_talk mailing 
list. For information on how to subscribe or unsubscribe, please visit 
http://www.chevreux.org/mira_mailinglists.html
[mira_talk] MIRA V3.2.1 available

Other related posts:
