Version 0.38

qtl2 0.38 (2025-06-02)

Minor changes

• In scan1() with a polygenic effect (that is, when kinship is provided) and when doing multi-core calculations, break up positions into intervals for more fine-grained use of the multiple cores.

• Added function compare_founder_geno() which is like compare_geno(), but comparing the founder genotypes.

Bug fixes

• Fixed seg fault that occurs in read_cross2() for multi-way RIL if cross info is misspecified or has missing values.

Version 0.36

qtl2 0.36 (2024-05-13)

Minor changes

• In scan1snps(), subset genoprobs and map to common positions,
  if they have different markers. (Issue #219)

Bug fixes

• Fixed a problem with sdp_panel=TRUE in plot_snpasso(). (Issue #232)

• Stop index_snps() with an error if physical map has missing
  values. (Issue #218)

Version 0.34

qtl2 0.34 (2023-11-28)

Major changes

• Changed read_csv() to fread_csv() to avoid conflict with readr::read_csv(). (Issue #223)

• Similarly, changed read_csv_numer() to fread_csv_numer().

• Added function find_dup_markers(), for identifying subsets of markers with identical genotype data. This is a port of qtl::findDupMarkers().

Minor changes

• Have calc_het() stop with an error if the genotypes have labels that aren't two characters, and add an explanation of this in the help info. (Issue #220)

• More fully explain the use of weights in est_herit() and scan1(). (Issue #221)

Bug fixes

• Deal with new compiler warning on CRAN. (Issue #230)

Version 0.32

qtl2 0.32 (2023-04-21)

Major changes

• In create_variant_query_func(), added new arguments id_field and sdp_field, and in create_gene_query_func(), added arguments name_field and strand_field (Issue #215). This gives new flexibility, but also adds new requirements (for example, that the variant database has a field "snp_id") and so could potentially break working code.

New features

• Added smooth_gmap() for smoothing out a genetic map, particularly to eliminate intervals with 0 recombination, by using a "mixture" of the map and constant recombination. Also added unsmooth_gmap() which does the reverse.

Minor changes

• read_cross2() now gives a warning if sex isn't provided but is needed. Also, if sex is missing we assume all individuals are female; previously we assumed they were male. (Issue #214)

• In plot_genes(), allow strand to be +/- 1 and not just "+" or "-". (Issue #216)

• Fixed date in citation, Broman et al. (2019) doi:10.1534/genetics.118.301595

Bug fixes

• In plot_genes() there was a case where stop was used that should have been end.

Version 0.30

qtl2 0.30 (2022-12-02)

Major changes

• For DOF1 and HSF1, revised the results of calc_genoprob() for the X chromosome, so that it just keeps track of the chromosome from the DO/HS parent. In males, we are assuming that the DO/HS parent is the mother.

Minor changes

• Added dependency on version of Rcpp (>= 1.0.7)

• Revised genoprob_to_alleleprob() to work with DOF1 and HSF1. plot_onegeno() should also work now in these cases. (Issue #140 and Issue #141)

Bug fixes

• Revised predict_snpgeno() to work for DOF1 and HSF1 populations.

• Now give a better error message in genoprob_to_snpprob() if snpinfo is missing the sdp column (Issue #207).

• In read_csv(), now give warnings if there are duplicate column names or duplicate row names in the file.

• In read_cross2(), moved the warning regarding the number of alleles to before the alleles object gets corrected (Issue #209).

• Now issue a warning message if founder genotypes are included but not used (Issue #211).

• The treatment of the male X chromosome in DOF1 and HSF1 was incorrect. We're now assuming that the DO or HS parent was the mother in the F1 cross, in which case males will be hemizygous for one of the DO/HS founder alleles.

Version 0.28

qtl2 0.28 (2021-10-11)

Major changes

• The default colors for the Collaborative Cross (CC) have been changed to a color-blind friendly palette. The original CC colors remain as CCorigcolors; the previous default is now CCaltcolors. The new colors are derived from the palette in Wong 2011 Nature Methods.

• plot_coefCC() was revised to include col=CCcolors as an argument. The default is the new color-blind friendly CC colors, but one can now more easily use col=CCaltcolors or col=CCorigcolors to get a different choice.

• Added plot_sdp() to plot the strain distribution patterns of SNPs using tracks of tick-marks for each founder strain. (Issue #163)

• Added arguments sdp_panel and strain_labels to plot_snpasso() so that you can include the plot_sdp() panel with the SNP association results and/or the genes.

Minor changes

• Added replace_ids() for a matrix or data frame (using the row names as the individual IDs). (Issue #191)

• Have calc_het() give an error if the input are for allele dosages. (Issue #190)

• Sneaky change in ind_ids() makes it apply to calc_genoprob and fst_genoprob objects. I'm not sure how to document this. (Issue #189)

• The output of est_herit() now includes the residual SD as an attribute, "resid_sd". (Issue #16)

• Implemented a cross type "hsf1" that is similar to "dof1", for a cross between an 8-way HS individual and a 9th strain. (Issue #149)

Bug fixes

• calc_kinship() died with cryptic error if genotype probabilities didn't have a names attribute; now using seq_len(probs).

• Give better error messages in est_map(), viterbi(), and sim_geno() if the cross is missing the genetic map.

• Fixed Issue #194: calc_genoprob() was taking chromosome names from cross$gmap which might have been missing; now using names(cross$geno).

• Fixed Issue #195: in create_snpinfo(), drop markers that are non-informative.

• Fixed Issue #196, that step() returns -Inf rather than NaN for general AIL. This had to do with the handling of -Inf in addlog().

• In fit1() and scan1coef(), wasn't grabbing the ... arguments. properly.

• Ugly c++ revisions to avoid clang UBsan warnings on CRAN. (Issue #169)

Version 0.24

Major changes

• Revised treatment of X chromosome in "general AIL" cross type, to be the same as the autosomes but with 2/3 as many generations for recombination. This should provide a better approximation.

Minor changes

• Revised reduce_markers() so that it can handle the case of many markers, by working with them in smaller batches.

• fit1() now returns both fitted values and residuals.

• fit1() can be run with genotype probabilities omitted, in which case an intercept column of 1's is used (Issue #151).

• Updated mouse gene database with 2020-09-07 data from MGI.

• Implemented Issue #184, to make calc_het() multi-core.

• Made the vdiffr package optional: only test the plots locally, and only if vdiffr is installed.

• calc_sdp() can now take a plain vector (Issue #142).

• Added a lodcolumn argument to maxlod() (Issue #137).

Bug fixes

• Fixed Issue #181, where calc_het() gave values > 1 when used with R/qtl2fst-based probabilities. Also fixed a similar bug in calc_geno_freq().

• Fixed Issue #172, where fit1() gave incorrect fitted values when kinship is provided, because they weren't "rotated back".

• fit1() no longer provides individual LOD scores (ind_lod) when kinship is used, as with the linear mixed model, the LOD score is not simply the sum of individual contributions.

• Fixed Issue #174 re genoprob_to_alleleprob() for general AIL crosses. We had not implemented the geno2allele_matrix() function.

• Fixed Issue #164, so plot_pxg() can handle a phenotype that is a single-column data frame.

• Fixed Issue #135, so plot_scan1() can take vector input (which is then converted to a single-column matrix).

• Fixed Issue #157, to have calc_genoprob() give a better error message about missing genetic map.

• Fixed Issue #178, to have read_cross2() give a warning not an error if incorrect number of alleles.

• Fixed Issue #180 re scan1() error if phenotypes' rownames have rownames.

• Fixed Issue #146, revising predict_snpgeno() so that it works for homozygous populations, like MAGIC lines or the Collaborative Cross.

• Fixed Issue #176, that guess_phase() doesn't work with cross type "genail". Needed to define phase_known_crosstype as "genailpk" in cross_genail.h because otherwise is_phase_known() will return TRUE.

Version 0.22-11

Additional small changes to fix problems on CRAN:

• Needed to remove some sqrt(int) that caused compiler errors on solaris
• Small fixes to get tests to pass on solaris
• Fixed problem with duplicate enums in c++ (because I don't really know what I'm doing)

Version 0.22-9

fixed compile error on solaris, on CRAN

• seemed to be just log(10) is not allowed; need to do log(10.0)
• also changed some 1 to 1.0 and a 2 to 2.0

Version 0.22-8

Numerous small changes to get the package on CRAN.

• Added R Core Team as a contributor, as the package includes a copy of code for Brent's method for univariate function optimization, taken from R version 3.2.2. Also added a Copyright field in the DESCRIPTION file.

• Sped up some of the examples and tests. Tests no longer use more than 2 cores (even those that are only run locally). A lot of back-and-forth about \dontrun{} and \donttest{}.

• Added \value{} sections in the documentation for various functions. Added further explanation of the "viterbi", "scan1", "scan1perm", etc. S3 classes in the documentation.