Skip to content

Latest commit

 

History

History
278 lines (173 loc) · 8.68 KB

README.md

File metadata and controls

278 lines (173 loc) · 8.68 KB

GCLr

Overview

GCLr is an R package designed to streamline genetic data analyses performed by Gene Conservation Lab (GCL) staff. Many of the functions in this package require data pulled directly from the GCL Oracle database Loki. Because of this, only Alaska Department of Fish and Game staff with credentials for accessing Loki will be able to use this package to its full extent. However, the package does allow for users without Loki credentials to read in genetic data contained in GENEPOP and rubias formatted files and convert the data into the objects used that can be used by GCLr (see GCLr::genepop2gcl() and GCLr::base2gcl).

Here are some examples of things that this package can be used for:

  • Laboratory workflow

    • pull data reports directly from Loki into R

      • get_asl_data() gets age, sex, and length data

      • get_extraction_info() gets DNA extraction information

      • get_geno() gets raw genotypes

      • get_gtseq_metadata gets GTseq metadata

      • get_tissue_data() gets genetic tissue data

      • get_tissue_locations() gets location of archived tissues

    • create lab project sample sheets

      • get_gtseq_sample_sheet() creates a GTseq project sample sheet

    • break up Loki import files

      • split_gtscore_loki_import() splits a GTscore import file into multiple files

  • Laboratory quality control (QC)

    • qc_template a template .RMD to compare the original (project) and reanalyzed (QC) sample genotypes to check for lab errors in data so they can be corrected, and calculates failure and error rates of genotypic data

  • Quality assurance (baseline and mixed stock analyses)

    • remove_ind_miss_loci() removes samples with missing genotypes

    • dupcheck_within_sillys() identifies duplicated samples

    • remove_dups() removes one sample from each duplicate set

    • find_alt_species() checks for wrong species

    • read_genepop_hwe() reads in GENEPOP Hardy-Weinberg equilibrium test results

  • Baseline analysis

    • Population structure

      • collections_map() creates an interactive map of collection locations

      • fishers_test() tests for homogeneity of allele frequencies,

      • pool_collections() combines collections into populations,

      • read_genepop_dis() reads in GENEPOP linkage disequilibrium test results

      • summarize_LD() summarizes GENEPOP linkage disequilibrium test results

      • create_pwfst_tree() creates a phylogenetic tree based on pairwise FST

      • create_mds_plot() creates an interactive multidimensional scaling

      • locus_stats() calculates observed heterozygosity, FIS, and FST by locus

    • Baseline evaluation

      • create_rubias_base_eval() creates rubias mixture and baseline objects/files for evaluating baseline reporting groups

      • run_rubias_base_eval() runs tests to evaluate the identifiability of baseline reporting units for genetic mixed stock analysis

      • plot_baseline_eval() plots the results of the baseline evaluation tests

    • Individual assignment (IA) evaluation

      • loo_rate_calc() calculates leave-one-out (LOO) error rates for each reporting group

      • plot_loo_prec_rec() plots LOO results as a precision-recall curve

      • IA_thresholds() calculates IA probablity thresholds based on precision-recall standards

  • Genetic mixed stock analysis (MSA)

    • create_rubias_base() creates a rubias reference (aka “baseline”) object/file

    • create_rubias_mix() creates a rubias mixture object/file

    • run_rubias_mix() runs an MSA in rubias

    • custom_comb_rubias_output() summarizes the rubias MSA results

    • stratified_estimator_rubias() combines estimates for multiple strata into a single set of estimates

    • summarize_rubias_individual_assign() summarizes the rubias individual assignment results

  • Data conversion

    • gcl2fstat() creates a genotypes file in FSTAT format

    • gcl2nexus() creates a genotypes file in NEXUS format

    • gcl2genepop() creates a genotypes file in GENEPOP format

    • genepop2colony() creates a genotypes file in COLONY format

    • genepop2gcl() reads in a GENEPOP file and creates .gcl objects

    • base2gcl() takes a rubias baseline object and creates .gcl objects

Installation

You can install the package from GitHub using pak.

install.packages("pak")

pak::pak("commfish/GCLr")

Reporting issues and requests

If you have any issues running the functions in this package, please file an issue on GitHub.

Issues can also be filed if you want to request enhancements to functions or additional functions to be added to the package.

Package maintenance

This package generally follows the git-flow branching model, using semantic versioning to document releases. Below is a quick summary of the different branches.

  • main

    • stable version of the package, commits/merges to main trigger a new version number

  • develop

    • ongoing, general improvements to the package including minor bug fixes

  • feature-branches

    • specific improvements to the package (i.e. creating functions for individual assignment)

  • hotfix

    • for fixing a serious bug found on main in the latest version of the package

    • references an issue

    • merge back into main and develop, triggers a new version number

Update version of GCLr package

Below is a generalized protocol for updating the package version by merging changes from the develop branch into the main branch. If working from a feature-branch, follow this same workflow for feature-branch –> develop, then develop –> main.

  1. commit changes on develop branch

  2. update NEWS.md, but without the version header #1, commit

  3. update the package version with usethis::use_version(), choose major, minor, or patch, have it commit for you

  4. push to develop

  5. create a pull request (base: main compare: develop) with meaningful title (i.e., merging to version 1.X.X) and brief description

  6. pull requests to main require review so we can keep main stable!

  7. merge after folks approve, confirm the merge, do not delete develop branch

  8. e-mail all GCL staff to notify everyone about the update

Hotfix protocol

  1. pull from main to make sure you have the latest and greatest version

  2. if your serious bug still exists, create an issue

  3. create hotfix_issue_XX branch (referencing the issue #) from main

  4. working on the hotfix_issue_XX branch, commit necessary changes to resolve the issue, note that you can include keywords in your commit that will auto-magically close the issue once the hotfix_issue_XX branch is merged back into main

  5. update NEWS.md, but without the version header #1, commit

  6. update the package version with usethis::use_version(), choose patch, have it commit for you

  7. push to hotfix_issue_XX

  8. create a pull request (base: main compare: hotfix_issue_XX) with meaningful title (i.e., Hotfix issue #) and brief description

  9. pull requests to main require review so we can keep main stable!

  10. merge after folks approve, confirm the merge, do not delete hotfix_issue_XX branch yet

  11. create another pull request (base: develop compare: hotfix_issue_XX) with meaningful title (i.e., Hotfix issue # merging to develop) and brief description

  12. merge, confirm the merge, now you can delete the hotfix_issue_XX branch

  13. e-mail all GCL staff to notify everyone about the update

Other helpful resources

ADF&G Division of Sport Fisheries Introduction to Git provides a good overview of Git, however, note that generally assumes that you will be working off of the shared network drive, rather than cloning to your local C:/ drive.

ADF&G’s Reproducible Research R Best Practices

GitKraken is a nice GUI alternative to GitHub for visualizing the commit/branch network