LoginSign Up
← Back to all sparks
O

onemap

INFRA · APIS
Velocity0.0

Construction of Genetic Maps in Experimental Crosses

github.com/cristianetaniguti/onemap ↗

A genetic-mapping mainstay that now points new users toward MAPpoly at load time

genetic-mappinglinkage-analysisgenotyping-by-sequencingmappolyinteroperability
◆Current state
OneMap constructs genetic linkage maps for experimental crosses, and its 3.x line turned it into a hub in a wider mapping toolchain: export functions for VIEWpoly and QTLpoly, summary and interactive ordering functions adapted from MAPpoly, and retention of reference and alternative allele information from imported VCFs. The most recent release, 3.2.0 in January 2025, adds an announcement about MAPpoly to the README and the package load message, fixes a subscript-out-of-bounds warning in rf_2pts and refreshes vignettes and tests.
◆Where it's heading
Versions 3.0.0 and 3.1.0 were driven by a benchmarking preprint on genotyping-by-sequencing best practice, and their content follows directly: marker filtering from two-point estimates, memory reduction after filtering, and optimisation of find_bins and map_avoid_unlinked. What has happened since is signposting rather than capability — a package that spent two releases integrating with MAPpoly now tells users about it every time it loads. The pre-3.0 tags are backfilled, with versions 2.3 through 2.7 sharing a single August 2021 timestamp, so their order carries no information.
◆Prediction
The load-message announcement suggests attention is flowing toward the MAPpoly side of the toolchain; expect maintenance and compatibility releases here rather than new mapping algorithms.

◆Recent moves

  1. 1y ago

    MAPpoly announcement added to README and load message

    Puts an announcement about MAPpoly in the README and the package startup message, alongside a fix for a subscript-out-of-bounds warning in rf_2pts and vignette updates. Adding a pointer to another package at load time is a deliberate act of direction-setting, and it is the only thing of note in the release.

    View source ↗
  2. 3y ago

    Exports to VIEWpoly and QTLpoly; marker ordering and map summaries

    Opens the package outward: export_viewpoly() and export_mappoly_genoprob() hand results to VIEWpoly and QTLpoly, ord_by_geno() orders markers by chromosome position, and summary and interactive ordering functions are adapted from MAPpoly. It also keeps reference and alternative allele information from imported VCFs, which downstream tools need.

    View source ↗
  3. 3y ago

    Two-point marker filtering and memory reduction

    Adds filter_2pts_gaps() and rf_snp_filter_onemap() for filtering markers on two-point estimates, plus keep_only_selected_mks() to release memory afterwards, and optimises find_bins() and map_avoid_unlinked(). The practical response to the benchmarking preprint cited in the notes, where filtering quality determined map quality.

    View source ↗
  4. 4y ago

    PSOCK and FORK parallelization choice; vcfR as an import

    Lets users choose between PSOCK and FORK cluster types and takes vcfR as an import. Small, but the parallelization choice matters on a package whose HMM steps are the slow part.

    View source ↗
  5. 5y ago

    hmm=FALSE option for ordering algorithms

    Adds an hmm=FALSE option so ordering algorithms can skip the hidden Markov model step, alongside vignette and workflow updates. One of five tags sharing a single backfilled timestamp, so its apparent position in the feed is meaningless.

    View source ↗
  6. 5y ago

    MDSMap dependency removed

    A one-line release dropping MDSMap from the dependencies. Part of the same backfilled batch of 2021 tags.

    View source ↗