v1.9 beta 4 and v2.0 alpha posted

[Christopher Chang]

In addition to bugfixes, 1.9 beta 4 includes:

* Support for up to ~65000 contigs.

* Allele count filters (--mac, --max-mac).

* More usable --pca variant weights (allele codes are now saved)

The 2.0 public alpha release includes:

* A new backward-compatible set of file formats which are roughly as efficient to perform computation on as PLINK 1 binary, while adding support for phase, dosage, and (soon) multiallelic data, and frequently taking much less space.  Oh, and no more involuntary REF/ALT allele swapping.

* C and Python libraries for working with the new genotype file format.

* Firth regression, along with a more computationally efficient "firth-fallback" mode.

* "--pca approx", a port of EIGENSOFT 6's fast PCA approximation.

* Support for linear algebra on matrices with >2^31 elements (so --pca is no longer limited to ~46000 samples, as long as your system has enough memory).

* KING-robust kinship coefficients (--make-king, --make-king-table, --king-cutoff).  These remain accurate when good population allele frequency estimates are unavailable.

* Proper support for dosages.  When .gen/.bgen files are imported, hardcalls *and* dosages are saved to the .pgen.  Operations which naturally extend to decimals (e.g. --pca, --glm, --freq, --maf/--mac) use the dosage information when it's present, while methods that can only make use of hardcalls (e.g. KING-robust, Hardy-Weinberg exact test) simply ignore the dosages. --hard-call-threshold can now be used to change the saved hardcalls without changing the dosages.

* Much more multithreaded code.

* Column sets which can be included/excluded from the main reports.

* Graffelman and Weir's extended chrX Hardy-Weinberg exact test, which takes male allele frequencies into account.  We've found that this tends to identify quite a few obviously miscalled chrX variants which were not caught by the usual QC filters.

* Import/export of Oxford-style haplotype filesets (--haps, "--export haps"/"--export hapslegend")

* Efficient export of sample-major PLINK 1 .bed files ("--export ind-major-bed")

Over the coming months, we will be backfilling many things PLINK 1.9 can do but 2.0 can't (first up: basic linear regression), but for now, don't expect to be able to use 2.0 without also having 1.9 around to fill in the gaps.