21 Jan: “–pull diversity” and “–exclude variety” don’t mistake out whenever its type in files incorporate a great chromosome password missing regarding current dataset.
: Fixed –meta-analysis-se-profession bug. –meta-analysis-chr-occupation and you can –meta-analysis-bp-community flags extra. On invention make, –dfam continuously problems out with the right message, instead of possibly segfaulting, when zero valid times exist.
twenty eight Oct: –meta-analysis-se-field banner extra, for interoperation having recent PLINK 2.0 logistic regression returns (which uses a ‘LOG(OR)_SE’ industry title so you’re able to disambiguate the product quality error’s tools).
Fixed “–model gen” permutation assessment insect
twenty-four Oct: Fixed uninitialized-varying bug when you look at the –dose + –q-score-assortment. “–indep-pairwise 500kb . ” really works securely once again (the earlier build called for a gap before ‘kb’).
cuatro Mar: Brackets inside command-range help text message are now actually used in a method a lot more similar to many other gadgets. –screening today performs properly whenever a beneficial numeric disagreement is step 1 earlier in the day the end.
twenty-six May (beta six): –merge-equal-pos bugfixes. For individuals who made use of –merge-equal-pos with .bed+.bim+.fam filesets where allele buy might have differed ranging from .bim files, or which have one .ped+.chart fileset anyway, i encourage redoing that run.
: –meta-study ‘report-all’ today reports unique stats rather than ‘NA’ when N=step one. Fixed an enthusiastic inaccuracy during the –adjust’s management of really small p-thinking. –make-just-fam really works securely with just .fam type in once again.
fourteen Late (beta 5): –file + –merge[-list] bugfix. chart file was unsorted, but when you have the opportunity we recommend redoing people –file + –merge and –document + –merge-listing multi-.ped merges performed having before plink 1.9 produces. (.bed consolidating are unchanged, and you may –merge-checklist instead –file was nearly never impacted.)
fifteen Oct: A2 allele is starting to become constantly reference towards the chr3 within a thousand Genomes phase step 1 downloads. Typically, do not expect you’ll keep REF/ALT alleles upright that have PLINK step 1.x, however, we tried to generate A2 continuously source within these data whilst still being didn’t some succeed; disappointed about this. (As to the reasons was that it discovered today, more than dos.five years after the truth? Just like the PLINK dos.0 is now offering an excellent –ref-from-fa flag that can easily seek such mistake.)
13 Oct: “–freq matters” no further gives stopped results when alleles have been flipped from the initial .bim/.bed buy. Fixed –logistic bug that could bring about spurious NA overall performance.
7 Oct: Fixed area-case (a couple of five haplotypes haven’t any known findings) bug in –ld/–r2/–blocks/–clump’s mathematical phasing techniques. 32-part make can now types >2GB .bed data files rather than crashing. –annotate and you can –gene-declaration today manage –allow-extra-chr.
fifteen Aug: –lasso is always to now work properly when there will be trials along with covariates expose nevertheless head phenotype shed.
21 Could possibly get: “–linear communications intercourse” and you can “–logistic correspondence intercourse” now work properly when zero covariates occur. –tucc set in creativity create.
20 March (beta cuatro): –pca var-wts now profile allele rules. Small statistical phasing bugfix. –[max-]mac computer, –R, and you can –thin-indiv[-count] not classified just like the unpredictable. PLINK dos.0 leader yields are now actually readily available.
It bug are much more likely to produce if the earliest
13 November: –geno/–shed now lose heterozygous haploid calls because nonmissing; this is consistent with –notice, and you may PLINK step 1.07’s choices that have –recode/–make-sleep productive. They might be today contradictory as to what PLINK step 1.07 did instead of –recode/–make-bed; have fun with –make-bed + –set-hh-destroyed basic if you’d like verwenden einen Weblink the old behavior. “–ci 0.95” no further produces ‘L94’/’U94’ column names whenever plink gets unlucky which have rounding. Slight –exclude/–extract/–flip bugfix to have type in records and no newline at the end. –qfam[-within] bugfix. –dosage ‘sex’ modifier today functions securely whenever particular trials was in fact blocked away (e.g. because the specific phenotypes was indeed destroyed).
20 September: VCF export no further removes an enthusiastic alt allele missing on quick dataset, on account of an excellent retroactive enhance with the VCF specs.