thirdParty/bgen/CHANGELOG.md

In weizhouUMICH/SAIGE: Efficiently controlling for case-control imbalance and sample relatedness in single-variant assoc tests (SAIGE) and controlling for sample relatedness in region-based assoc tests in large cohorts and biobanks (SAIGE-GENE)

History

7 July 2016

• Updates to bgenix to handle UK biobank interim files and to avoid extra index tables in the index file.

21 March 2016

• BGEN spec and implementation updated to alter probability order for unphased data when the number of alleles (K) or the ploidy is greater than two. This order now better matches the order of VCF GP fields and as a simple enumeration scheme.

10 Nov 2015

Major changes in revision ff11254f9505:

1. I've implemented two new tools

   • cat-bgen, which can be used to concatenate BGEN files.
   • bgenix, which can be used to index BGEN files and efficiently retrieve specified data.

2. For this purpose I've imported several extra pieces of code

   • appcontext/ and db/ sublibs from qctool
   • sqlite3 3.9.2
   • boost 1.55.0

Note: these changes were erroneously applied first to the master branch (they were intended for default first).

6 Nov 2015

Major changes in revision 392429affc42:

1. I’ve changed the behaviour of BGEN v1.2 with respect to samples with missing data: they are now stored with dummy zero probabilities. The spec is now in 'beta' which means I don’t have any other planned changes to make; unless major issues are uncovered this will be the final version of the format.

2. I’ve revamped the setter api of parse_probability_data somewhat. It is documented in the code and here on the wiki. The main breaking changes are:

3. Renamed operator() to set_value(), and given it an index argument; I think these make the API more consistent.
4. Added an initial ploidy argument to set_number_of_entries() as requested. (The type of data - phased or unphased - is already reported in the order_type argument so I don’t think another argument is needed).

5. Added two new method calls, which are optional: set_min_max_ploidy() (useful for setting storage) and finalise(). See the docs for info.

6. I’ve also got rid of the max_id_size option to write_snp_identifying_data(). (This is now not needed because writing BGEN v1.0 files is no longer supported.)

7. I’ve also added some test code (using the catch framework, which seems pretty good). Tests are not exhaustive but hopefully a start. :q

8. I've removed some code warnings - thanks to Robert V. Baron of Mega2 for testing this code.

23 Sep 2015

First version, based on qctool implementation.

weizhouUMICH/SAIGE documentation built on May 6, 2022, 12:34 a.m.