Welcome, Guest. Please Login
YaBB - Yet another Bulletin Board
 
  HomeHelpSearchLogin  
 
[ADDITION] Multiple Enhancements (Read 1530 times)
Sergei
YaBB Administrator
*****
Offline


Datamonkeys are forever...

Posts: 1658
UCSD
Gender: male
[ADDITION] Multiple Enhancements
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
Back to top
 

Associate Professor
Division of Infectious Diseases
Division of Biomedical Informatics
School of Medicine
University of California San Diego
WWW WWW  
IP Logged