transrate: v1.0.0 final
| Main Authors: | Richard Smith-Unna, Chris Boursnell, The Gitter Badger |
|---|---|
| Format: | info software Journal |
| Terbitan: |
, 2015
|
| Online Access: |
https://zenodo.org/record/18584 |
Daftar Isi:
- transrate v1.0.0 final This is the final release of transrate v1. This our first stable production release. Please report any bugs on the issue tracker. Documentation for this release is on the transrate website: http://hibberdlab.com/transrate/. Installation Simple install This is the simplest option if you just want to use transrate from the command-line... Download one of the binaries linked below, unpack it, and add it to your PATH. Linux 64 bit: transrate-1.0.0-linux-x86_64.tar.gz OSX: transrate-1.0.0-osx.tar.gz Download the example data so you can test your installation has worked. Advanced/developer install This is the option you need if you want to develop code that uses transrate as a library... $ gem install transrate $ transrate --install-deps Changes If you are updating from a version below v1, please see the release notes for v1.0.0.alpha1. Changelog v1.0.0 final All metrics are now reported to an appropriate number of significant figures Our dependency on SNAP has been updated to v1.0beta.18 after they incorporated bug fixes we submitted Our dependency on Salmon has been updated to v0.4.2 to include some bug fixes The command-line interface has been cleaned up and simplified A bug with dynamic loading of C extensions, introduced in a previous commit, has been fixed beta4 We've updated our Salmon dependency to v0.4.0, which brings an awesome speed increase, especially for small datasets, as well as some improvement in accuracy for our read metrics. Fixed a bug in --install-deps where it would always try to install all the dependencies Improved handling of FASTA input edge-cases and unusual formatting Better handling and reporting of errors and crashes in the dependencies A big improvement to reference-based metrics when comparing to a reference that has repetitive sequences. Users should now see considerably higher reference coverage in some places (before, BLAST was masking out repetitive regions). Thanks to Nick Schurch and Chris Cole at Dundee for their analysis that led to discovering this issue. Some command-line arguments have been reordered and their short-names changed, so now --right has -r for short and --reference has -a for short. Thanks to Matt Macmanes for highlighting this. beta3 Transrate will automatically choose the optimal set of contigs from an assembly, maximising both the quality of the contigs and the proportion of the read evidence that is used. It now reports the raw assembly score as well as the optimised assembly score. The optimal assembly is found in good.*.fa Transrate can now merge assemblies with the --merge-assemblies command, choosing the optimal set of contigs from all the input assemblies Some great improvements to Salmon have been included in this release Transrate now estimates the insert size distribution from the reads Transrate is now distributed as a binary package (available for linux 64bit, and osx). This should greatly simplify installation for most users. beta2 We have moved to using Salmon for read assignment. This has brought a massive speedup over using eXpress, because Salmon can use any number of cores (while eXpress is restricted to 3), and Salmon is generally optimised for speed We've managed to eliminate much of the IO bottleneck that was caused by processing BAM files, thanks to the collaboration of the Salmon developer We've fixed some bugs in the SNAP aligner that we depend on Minor bug fixes in basic contig metrics for unusual kinds of inputs Improved error handling and logging Improved dependency management, with fewer dependencies Acknowledgements Rob Patro, developer of Salmon, for his hard work on making Salmon work well with transrate Users, especially Matt Macmanes, for diligent testing and bug reporting The SNAP team for responding rapidly to our bug reports