September Changes in BioC Devel

    September 2005

aCGH
    Reimplemented hmm from scratch
    Removed dependence on repeated and rmutil packages
    Removed as.eSet function
    Changed initial diagonal probability of the transition probability matrix to 0.99
    Set the posterior probabilities to be used for picking the optimal decoding
    Added Makevars to fix Mac compilation problem

affxparser
    Fixed read.cdf.env to transpose the PM/MM matrices
    Added PM, MM column names

affy
    Fixed a typo in justRMA man page
    Removed dependence on reposTools
    Changed cleancdfname so multiple calls do not repeatedly add "cdf"
   
affycoretools
    New package:
        Various wrapper functions that have been written to streamline the more common
        analyses that a core Biostatistician might see.

    Maintainer:
        James W. MacDonald

affyPLM   
    Fixed a possible garbage collection problem in threestepPLM
    Fixed threestepPLM so it works on Windows
   
AnnBuilder
    UniGene URL fixed
    Added two arguments "title" and "url" to ABPkgBuilder
    Corrected some bugs that cause warnings in R CMD check
    Fixed a bug for probeset IDs containing a "#"
    Added new environments PFAM and PROSITE from IPI website
    Fixed bug reported by Li Long in cMAP
    Fixed bug in writeManPage
       
annotate
    Deprecated installDataPackage()
    Removed dependence on reposTools

Biobase
    Changes for eSet revisions
        reporterInfo slot in annotatedDataset
        data.frameOrNULL union to support legacy annotatedDataset objects
        Added swirl data for eSet vignette
    Removed dependence on golubMerge
    Removed dependence on hu6800
    Version update to 1.7.0
        Major code reorganization - no changes in functionality
        Updated man pages - added Biobase-package and simplified package index
    Replaced class of description slot in eSet from characterOrMIAME to MIAME
    Removed dependence on reposTools
    Added biocReposList, a function to get a character vector of BioC repositories that are
      appropriate for that version of Biobase. Used to remove dependence on reposTools for
      the automagic downloading of packages by e.g., the affy package

biomaRt
    Added first version of BioMart API functions
    Function bmget replaced by getBM
   
DNAcopy
    Changes to the plot command allowing user to control plotting character, color and x-axis
    Changed the plot function so par("ask") behaves better
    Fixed bug that caused segfault when window.size is set if p.method is hybrid
    Added options for zeroline, line weight, etc
    Fixed ylim bug that calculated the limits from the first sample instead of all samples
   
gcrma
    Restored default fast = TRUE
    Made justGCRMA compatible with gcrma
   
geneplotter
    Fixed bug due to a change in the API for axis()
   
gff3Plotter
    Most of the code rewritten from scratch
        Simplified function definitions
        Removed dependencies
        Improved graphics and color scheme
    Variable initialization corrected in gffPlotting
   
GlobalAncova
    Corrected vignette
    Corrected horizontal.bars
   
goCluster
    Fixed bug in configuration dialog
   
GOstats
    Fixed example for shortestPath
   
graph
    First pass at node attributes implementation
    Fixed interface for vectorized assignment for node attributes
    Added notes on refactored interface to graphAttributes vignette
    Replace edgeSet and nodeSet with attrData backing store
    Added edgeData and nodeData slots to the graph virtual class
      This gives generic attributes to all graph subclasses
    Rename graphIM to graphAM to coincide with adjacency matrix in monograph
    Documentation fixes and additions
    Added coerce method graphAM --> graphNEL

ideogram
    Moved buildChromLocation.2 from reb package
   
iSNetwork
    Added functions to create chilren models
    Fixed bug in the callback functions
    Finished the interface for creating a GO graph child model
    Added some functions to convert messages between graphs and other models   
    Created a user interface for creating a child GO graph model
    Removed unused functions
    Added functions to create a GO graph from a set of Affy IDs
   
   
limma
    Updated man page for readImaGeneHeader to match function
    Changed multiplicity adjustment to "BH" for topTable, decideTests, etc.
    Updated to 2.0.11 - updated usersguide.pdf only
   
makePlatformDesign
    Added functions to check NimbleGen NDF compliance
    Added functionality to use 10K Affymetrix SNP chips
   
maSigPro
    New package:
        maSigPro is a regression based approach to find genes for which there are significant gene expression profile
        differences between experimental groups in time course microarray experiments.

    Maintainer:
        Ana Conesa

MVCClass
    Added functions to create children models
    Fixed bug in the callback functions
    Added functions to convert messages between graphs and other models
    Removed unused functions
    Added functions to create a GO graph from a set of Affy IDs
   
oligo
    Added functions to check NimbleGen NDF compliance
    Removed dependence on reposTools

pdmclass
    Added a vignette
    Fixed my.svd man page
    Added missing man pages
   
prada
    Added cytometry gating functions
    Added threePanelPlot function
    Fixed example in plotPlate
   
RBGL
    Fixed brandes.betweenness centrality
    Hyperlinks to all implemented interfaces
    Updated info in 'See Also' portion of man pages
    Added info for betweenness.centrality.clustering

reb
    Added buildChromLocation.2
    Updated summarizeByRegion man page
    Updated mcr.eset to conform to eSet
    Moved buildChromLocation.2 to ideogram package
   
reposTools
    Deprecated develOK option to install.packages2 - devel packages now require devel version of R
    Note: reposTools will be deprecated in BioC 1.8
   
Rgraphviz
    Disable fdp layout on Windows
    Updated for graphviz 2.6

RLMM
    New package:
        A classification algorithm, based on a multi-chip, multi-SNP approach for Affymetrix SNP arrays.
        Using a large training sample where the genotype labels are known, this aglorithm will obtain more accurate classification results on new data.
        RLMM is based on a robust, linear model and uses the Mahalanobis distance for classification.
        The chip-to-chip non-biological variation is removed through normalization. This model-based algorithm captures the similarities across genotype groups and probes,
        as well as thousands other SNPs for accurate classification. NOTE: 100K-Xba only for now.

    Maintainer:
        Nusrat Rabbee

tilingArray
    Updated plot.segmentation to handle large segmentation objects
    Updated naming of hybesets
    Added setScriptsDir, which sets the directory from which helper scripts are sourced
    Added lowTranscribedRegions to find apparently low transcribed regions for PCR verification
    Fixed threshold calculations
    Modified alongChromosome plots to let features be annotated by specified order
    Added clustering of the windows to lowTranscribedRegions
    Fixed readSegments
    Adapted tableSegments and categorizeSegments to new annotation
    Better UTR map / GO analysis
    Nicer plots
    writeProbeSeqsFasta writes a FASTA file with the probe sequences
    New analysis of relationship between presence of antisense transcript and UTR lengths
    Fixed antisense-GO to use both total and poly-A
    GOHyperG now prints gene lists only if length <= 36
    UTR/antisense association analysis now uses both poly-A and total