- Switch MungeSumstats from GitHub remote to Bioconductor dependency, fixing installation failures when GitHub PAT is unavailable.
- Replace corrupted BST1 VCF with fresh 1KG Phase 3 subset.
- Fix Rd formatting and
\dontrun{}example guards. - Fix
test-get_headerVCF column count assertion to match replacement VCF. - Fix
test-previewto check column names in output instead of brittlefread(text=...)round-trip that breaks across data.table versions. - Update install instructions to use
BiocManager::install(). - Fix Codecov icicle graph URL (
codecov.ioinstead ofapp.codecov.io). - Add
Rplots.pdfand*.tbito.gitignore.
- Local R CMD check fixes and compatibility updates.
- Add
docker_registry: ghcr.iotorworkflows.ymlfor GHCR Docker builds.
melt_finemapping_results:- Check to see if already melted.
- Update rworkflows.yml
portal_query:- Ensure
files_filtis indata.tableformat.
- Ensure
reassign_lead_snps- Ensure
merged_datis indata.tableformat.
- Ensure
- parquet functions:
- Remove unit tests for now, as they cause segfault errors during checks (but not when run manually).
- Use
rworkflowsGHA. - Moved all GitHub-related functions to new package:
echogithub- Kept portal-related function in
echodata: e.g.portal_query
- Kept portal-related function in
- use remote
MungeSumstatsuntil it passes bioc devel again
import_topSNPs:- Automatically downloads data is
topSSis a remote file. - New arg
startRowfor skipping rows.
- Automatically downloads data is
check_if_empty-->is_empty(more concise)get_data:- Allow users to specify
tagand ... args.
- Allow users to specify
get_headercan read parquet now.
standardize_maf:- Fix flaw in logic preventing MAF being imputed from Freq.
standardize_proportion_cases:- When
colmap$proportion_casesis NULL, just returnquery.
- When
- Fix GHA: @master --> @v2
- Fix
reassign_lead_snpswarnings. - Fix
biomart_geneInfowarnings. - Fix
is_urlexample. gene_locus_list:- Drop loci missing from topSNPs.
- Fix
melt_finemapping_results:- Alphanumerically sort
Methodscol consistently via new internal function:order_methods
- Alphanumerically sort
get_sample_size:- Add more complete documentation on what happens when "N" column is present.
- Add extra check to make sure
MungeSumstats>=1.3.9 is installed. compute_ncan now be a numeric vector of length equal tonrow(dat).
is_url:- New export function.
- User-specified URL
protocolsto search for.
- Make
R.utilsan Import to avoid any issues with reading gz files. - Make sure
standardizeusesconstruct_colmapto rename "N" column.
- New function:
get_Kunkle2019_coloc. - Remove
zscore*functions (not used anymore). standardize_gene_locus_cols: Will replaced Locus names with new Locus_Gene naming scheme whengrouping_varscontain both Locus and Gene.- Move over functions from
echoannot:get_SNPgroup_countsresults_reportmerge_finemapping_results
portal_query: add arg to return results in table format.- Export
get_data get_Nalls2019_loci:- Add
limit_snpsarg = Savetempdir()instead of cache by default, to allow subsetting different numbers of SNPs more easily. - Add
force_newarg.
- Add
portal_query: Forgot to addsep=name in URL construction.get_sample_size: Fix unit tests.
get_header: Fix misnamed variable.
- Pass up
compute_nas arg instandardize. - Add
get_Nalls2019_colocfunction.
get_sample_size: Return NULL when "N" col not present at end of func. Otherwise,max(NULL)can give you-Inf.
- Remove
MungeSumstatsfrom Remotes. - Passing all CRAN checks.
- Cache datasets.
- New
get_Nalls2019_locifunction, to easily write example loci to disk and return paths.
- Use
pasteinstead offile.pathto construct portal query URLs.
get_header: Can now handle VCFs (assuming #CHROM is the comment char).- Moved
get_CS_binsandget_CS_countsfromechoannot-->echodata. - Passing all CRAN checks!
- Fix GHA pkgdown building:
- The newest version of git introduced bugs when building pkgdown sites from within Docker containers (e.g. via my Linux GHA workflow). Adjusting GHA to fix this.
construct_colmap: Collapses many column mapping args into one function.standardize_subset-->standardize:- Simplified workflow.
- Operates on only
queryinstead of bothqueryandquery_mode. - Created subfunctions for each step.
- Removed
liftoverstep.
- New functions:
previewis_localis_url
- Increased test coverage to >88%.
- Removed
standardise_sumstats_column_headers_crossplatformto use directly fromMungeSumstats.
import_topSNPs:- Now uses
standardise_sumstats_column_headers_crossplatformexported function whenmunge=TRUE. - Both standardisation strategies no longer interfere with one another when used at the same time.
- Now uses
- Added .zzz file with
.datatable.aware <- TRUE - Added new function
merge_robustto address issues withdata.table.- See here for Issues: RajLabMSSM/echolocatoR#72 (comment)
import_topSNPs: Added example and documented all params.- Rename
MUNGESUMSTATS.*functions tomungesumstats_. get_sample_size: Add unit tests and newreturn_onlyfunction for min/max.
- Added new functions (with unit tests):
snp_group_colorDictread_parquetwrite_parquet
reassign_lead_snps: Handle situations where none of thegrouping_varsare available.- Fix error in reporting number of Consensus_SNPs in
find_consensus_snps(only affected console message, not the actual data).
- Added new functions (with unit tests):
snp_group_filtersfind_consensus_SNPsfind_consensus_SNPs_no_PolyFunfilter_snpslimit_SNPsgene_trimmerfillNA_CS_PPdetermine_chrom_typeis_emptycolumn_dictionaryget_headermelt_finemapping_resultsupdate_colsbiomart_geneInfofind_top_consensusdt_to_grangesis_grangesis_ggplotis_ggbiogranges_to_bedreassign_lead_snps
- Added new function
coords_to_rsids, copied over fromcatalogueR.
- Added a
NEWS.mdfile to track changes to the package. - Added new function
assign_lead_SNPwith unit tests. - Added unit tests for
createDTandMUNGESUMSTATS.*functions. - Removed unused functions.
- Added dynamic badges to README.
- Updated GHA workflow.
- Added GH token to Secrets.
- Large datasets uploaded/downloaded via
piggyback.
- Functions extracted from
echolocatoRto createechodata.