HyPhy message board
http://www.hyphy.org/cgi-bin/hyphy_forums/YaBB.pl
HYPHY Package >> Feature Additions >> [ADDITION] Multiple Enhancements
http://www.hyphy.org/cgi-bin/hyphy_forums/YaBB.pl?num=1156262186

Message started by Sergei on Aug 22nd, 2006 at 8:56am

Title: [ADDITION] Multiple Enhancements
Post by Sergei on Aug 22nd, 2006 at 8:56am
I haven't really been posting the information updates with much regularity, but here are some worth mentioning:

  • Expanded compartmentalization analyses in Standard Analyses (e.g. Simmond's AI, and tree correlation coefficients)
  • A new set of analyses to estimate the distribution of rates in coding data (dN and dS) using a general bivariate distribution, i.e. allowing for dependence of dN and dS at a site at the level of a distribution (under Codon Selection Analyses:dNdSBivariateRateAnalysis.bf). This can also test for evidence of dN>dS, whether or not dS is variable, and whether or not HKY85 is appropriate to correct for nucleotide substitution biases.
  • A few data file tools: e.g. translate codon to protein (in batch language, suitable for automated pipelines), in Data File Tools. A simple codon based alignment tool (suitable for HIV and star-like phylogenies), including reference HXB2 genes if desired (SeqAlignment.bf)
  • Count and tabulate synonymous and non-synonymous ambiguities (suitable e.g. for bulk HIV-1 sequences): Miscellaneous: SandNSAmbigs.bf
  • AAModelComparison.bf: a tool to fit a number of protein models, including REV and REV restricted to amino-acids which are one nucleotide substitution away, and some in-house HIV models developed together with David Nickle of UW.
  • DatedTipsMolecularClock.bf. A spiffed up version of Andrew Rambaut's TipDate, which allows for nuc/aa/codon data and can test for dated clocks only on parts of model rates (e.g. only synonymous rates)
  • A number of bug fixes and internal feature additions, most notably a large library of context-free grammar definition and training (as far as we can tell, we may have one of the fastest and memory-miserly implementations out there), written mostly by Art Poon (for documentation and examples e-mail me). The abilitiy to serialize likelihood functions which depend on multiple data alignments. A feature to automatically parallelize single likelihood function calculations accross an MPI cluster (AUTO_PARALLELIZE_MPI = 1; flag) and many more


I'll try to be more regular with these updates...

Cheers,
Sergei

HyPhy message board » Powered by YaBB 2.5.2!
YaBB Forum Software © 2000-2024. All Rights Reserved.