Adding locusbreaker #96
Open
+66,711
−299
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
…landers R package
…n until merge with github flanders r repo
…age (temporary solution) and fixing parameters
… re-finemmaped with L=1. And overall L=1 loci do not go thorugh QC
…ng wiped out by post susie QC
…ject in a function
…d, default is fine
… parameters to required list
Update susie QC logic and allow for customizable parameters
…name to tiledb_bfile parameter in test
Collaborator
|
Thank you very much @bruno-ariano!!! I think it makes sense to keep the "summary statistics" and "TileDB + summary statistics" versions separated for the time being, while we test the TileDB version more extensively. Would it be possible to avoid publishing the |
Add credible set expansion, logsum and wrapping up susie reformatting in a function
…rmatting in a function"
…nsion Revert "Add credible set expansion, logsum and wrapping up susie reformatting in a function"
* Removing susie.cs.ht() because it should be taken directly from the flanders R package * Have customizable post susie QC parameters * Nevermind - function copied and pasted from gitlab, temporary solution until merge with github flanders r repo * Back to sourcing function rather than calling it from flanders R package (temporary solution) and fixing parameters * Forgot to remove susie_qc_cs_lbf_thr parameter from here * Added susie QC parameters to the nextflow schema * Rearraning post susie QC, so that loci disappearing becuase of QC are re-finemmaped with L=1. And overall L=1 loci do not go thorugh QC * Adding locusbreaker * Adding locus size parameter * Add report collecting loci that were re-finemapped with L=1 after being wiped out by post susie QC * Wrapped all code to go from susie output to rds list of dataframes object in a function * Adding credible set expansion and using function from susie output to rds format * Remove some parameters from list of those mandatory - if not specified, default is fine * Do not hardcode L=10, but rather use the assigned variable * Ok nevermind, reverting previous commit and adding also post susie QC parameters to required list * Removing hardcoding of L=10 for easier maintenance * Forgot to close parenthesis * Temporarely copy and paste functions from gitlab R package version * Computing also cs logsum - and adding it to the anndata obs * Fixed unmatching parenthesis * tile_lb_input parameters defined but not used. Assigning correct filename to tiledb_bfile parameter in test * Removing tuple since it's only one element * Revert "Merge branch 'tiledb_locusbreaker' into cs_expansion" This reverts commit fc15e3b, reversing changes made to 2059698. --------- Co-authored-by: arianna.landini <arianna.landini@external.fht.org> Co-authored-by: bruno-ariano <bruno.ariano.87@gmail.com>
* Replacing quit with next if no susie LD matrix is empty - avoid breaking the fine-mapping loop * Checking for KL length in loci with less than L SNPs
…to annData (#108) * Replacing functions - now going from susie output directly to anndata rather than to .rds * No longer saving .rds files but anndata (one per fine-mapping job) * Renaming logsum_lABF to logsum.logABF * Opt argument is batch, not batch_index * updating metadata tiledb * Replacing with grch38 and adding path, replace txt to csv extension in metadata * Adding input batching logic to tiledb metadata * Replacing txt with csv metdata extension, adding pgen version of grch38 ld * Adding batch size in tiledb test, using grch37 pgen version for ld * Adding resources for multi-cpus process * Use params rather than hardcoded batch size. Remove all views * Add batch index to input, set optional output, add batch-name argument * Update to TileDB fragment and metadata - gwas specific pvalues thresholds * Pval thresholds now taken from metadata * Removing view of channels * Not removiong study_id and phenotype_id columns - needed later for coloc - and adding chr to chromosome number * Adapting to updated column names in locus brekaer * Making concat_anndata more flexible - can be used to concatenate also output finemapping anndata, removing no longer needed rds output * Specifying running coloc in tiledb test * Removing no longer needed .rds output, adding reticulate to have anndata working * Still keep coloc master table output * Taking pheno variance calculated from TileDB --------- Co-authored-by: arianna.landini <arianna.landini@external.fht.org> Co-authored-by: bruno.ariano <bruno.ariano.87@gmail.com>
…ta to QCed finemapping only after checking the object isn't null. Now computing conditional statistics inside from_susie_to_anndata()
…temporarly switch off path_tiledb_lb_out
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This branch created the option of using Locusbreaker from TileDB. In order to do so I had to add a new module trying to leave as much as possible the following finemapping and the other munging_locusbreaker option.
Note that TileDB data was already munged and QCed.
I compared the finemapping and coloc results when running Locusbreaker from TileDB and from the original Flanders and the results are exactly the same