NEWS file for the pedgene package

Release Notes for version 4.1

• remove depencency on kinship2, replace with bioconductor Pedixplorer

• index rows of ped/geno by famid_id to match kinship dimnames, which makes robust to lc_collate char-sorting settings

Release Notes for version 3.7

• rename format.relation as format_relation

Release Notes for version 3.3

• Allow either ped or famid as the pedigree id in ped and geno datasets

• Allow special relationships for family members, as in kinship2 package

Release Notes for version 2.9

• Use trait.adjusted as mean of trait for those with genotype data

Release Notes for version 2.6

• Make sure character ids stay as character ids

• Bug fix for no relateds

• For X chrom, add check for monomorphic to include within males and females only

Release Notes for version 1.4

• Add argument acc.davies for pedgene. Some kernel tests gave NA p-values with default daves accuracy of 1e-6. Performs well with default of 1e-5, but give option to user in case NAs occur.

• return n-variants for the gene tests, now a column in the returned object from pedgene

• Allow the method to complete if any gene has no markers even after all are removed for having zero variance (same genotype)

• When a gene has only one marker, the kernel test reduces to burden test. Return same test stat and p-value for both. Thanks to Seth Ament.

• Fix bug where failed to subset genotype matrix when either NA in trait or all NA genotypes for a given person, after matching geno and ped subject by ped-person ids. Thanks again to Seth Ament.

• Change maintainer to Jason, rather than Dan.

• Add n.noninform to pedgene object, the number of non-informative markers that were removed

Release Notes for version 1.0

• Require ped and person ID columns in genotype data. To match to more-complete subjects in pedigree data by ped and person ID.

• Added argument checkpeds, allows user to turn off the step that performs pedigree checks.

• Kinship function (from kinship2) is now called without making a pedigree object first. This step allows disconnected pedigrees, and identifiers can be either numberic or character.

• Perform Davies' p-value for Kernel statistic.

• Added package vignette, a simple manual

Release Notes for version 0.5

• wrapper called pedgene to call pedgene.stats on one gene at a time while re-using kinship matrix for all pedigrees

• require ped and geno data.frames to have same number of rows

• add test script: test.pedgene.R

• if map given (with chrom and gene columns), require map to have as many rows as geno has columns

• define S3 pedgene object, with simple print and plot methods

Release Notes for version 0.1

• Original release, mostly code from Dan's scripts

• pedgene.init pre-calculates some stats, passed to pedgene.stats for one single gene