Malformed data files from specific old merges no longer bring about segfaults

dos Sep: –serving ‘noheader’ is to now work properly when specific phenotypes are lost. –vcf + –vcf-filter today parses semicolon-delimited Filter sphere truthfully. –vcf-min-gq flag placed into innovation build. –exclude/–extract not possess terrible results whenever both the fundamental dataset and the –exclude/–extract variant checklist have an incredible number of ‘.’ entries.

28 August: –recode HV[-1chr] today always uses ‘0’ as lost phenotype code, as Haploview does not deal with -nine. –write-covar + –with-phenotype no longer segfaults towards circumstances/manage phenotypes, and you will multichar allele .ped loader not any longer segfaults into the nonstandard tri/quadallelic alternatives. –meta-analysis put into advancement make.

14 August: Repaired a good typo from the eleven August build and therefore triggered covariate packing and some other features to hang.

11 August: –linear/–logistic no longer uses a buggy Huber-Light important mistake estimator whenever groups are discussed. –output-chr today really works safely which have –make-bed when the enter in .bim is unsorted, and a very long allele password inside the an enthusiastic unsorted .bim don’t causes –make-sleep to segfault.

step one August: –serving logistic regression bugfix. –make-just-bim and you will –make-just-fam flags extra. .bim/.fam data files may now become processed without an accompanying .sleep under certain situations. –recode-allele now performs properly which have A-transpose means.

18 July: Text message filesets having both multi-character allele codes and an unsorted .chart document no longer cause the autoconverter so you can crash. –pfilter is to now constantly filter ‘NA’ entries. –dose chromosome code returns insect repaired. –read-freq now plenty A1 allele requirements if they are destroyed regarding chief dataset, unlike erroring in that condition. –show-labels set in advancement generate.

Very first Cochran-Mantel-Haenszel and Breslow-Time assessment set in innovation generate

cuatro July: Fixed a couple sitios de citas hindú of merge pests and therefore potentially caused investigation in the last pair examples of an input otherwise merged fileset as mishandled. 90 alpha/beta yields, and most two samples had been inside, i suggest redoing the brand new mix that have a more recent create. (Extremely merges was unchanged, however, most readily useful safe than simply sorry.)

For those who have blended filesets produced with prior to PLINK step 1

step one July: Boost has just put (disappointed about that) –data/–gen/–test command-range parsing bug. –indep no longer misreports what amount of pruned variants whenever truth be told there are detailed multicollinearity. –fst, –homog, and you can –oxford-pheno-term flags added.

twenty-seven June: “–recode oxford” not dumps incorrect IDs whenever found in a similar work on while the an example filter. –fill-missing-a2 flag added. –recode ‘A-transpose’ and ‘include-alt’ modifiers added. “–het short-sample” means added.

20 Summer: Fixed –mendel no chromosome code/segfault insect. –merge-listing no more requires a research fileset. –fast-epistasis + –parallel bugfix. –indep-pairphase and QFAM try completed in development generate.

ten Summer: –merge-mode 1 now truthfully merges destroyed calls that have one nonmissing label. –r/–r2 chromosome boundary dealing with bugfixes. ‘0X’/’0Y’/’0M’ chromosome requirements emitted by Oxford equipment are in reality acknowledged, and then have backed by –output-chr. –vcf-half-name flag set in control management of ‘0/.’ VCF GT viewpoints, and standard choices is ‘error’ function to force a mindful ent build.

5 June: –ld-snp-record and you can multipass –r/–r2 bugfixes. Nonstandard ‘0/.’ and you may similar VCF GT field opinions are actually canned just like the if they didn’t have the brand new at the rear of ‘/.’, in lieu of causing a segfault. (Handling of this example may be configurable down the road, tune in.) –linear/–logistic permutation bugfix (permutation achievements count variety was not initialized to all the-zero). –dosage linear regression set in innovation make.

step three Summer: Gzipper no longer deletes new yields file whenever becoming expected to help you append (it was ultimately causing larger n-violation computations including –r2 gz to only keep the production of your own past admission). Fixed –genome twice-missing-label addressing insect, and you will a beneficial –r/–r2 blended autosomal/nonautosomal data handling bug. “–dosage exist” and –write-dose set in innovation create.

e-related segfaults. “–r2 dprime” missing data handling bugfix. Corrected misnamed –filter-attrib[-indiv] flags to –attrib[-indiv], fixed a positive matching bug, and added support for gzipped attribute files. “–recode tab” don’t gives off places throughout the heading line. –stops is now offering a good ‘no-pheno-req’ modifier and that eliminates the excessive phenotype requirements. –annotate put in advancement create.