dos Sep: –dose ‘noheader’ should now work effectively when some phenotypes are forgotten. –vcf + –vcf-filter today parses semicolon-delimited Filter fields correctly. –vcf-min-gq christliche Liebe Dating-Seite flag put into innovation build. –exclude/–extract no longer features terrible results whenever both chief dataset in addition to –exclude/–pull variant number has actually many ‘.’ records.
28 August: –recode HV[-1chr] today always uses ‘0’ as the lost phenotype password, just like the Haploview does not deal with -nine. –write-covar + –with-phenotype not segfaults towards instance/handle phenotypes, and multichar allele .ped loader no more segfaults to the nonstandard tri/quadallelic variations. –meta-data added to creativity make.
fourteen August: Fixed good typo regarding eleven August build and therefore brought about covariate loading and some other qualities to hold.
eleven August: –linear/–logistic not spends a buggy Huber-Light basic error estimator whenever groups is actually outlined. –output-chr today works safely with –make-bed when the enter in .bim is actually unsorted, and you will a long allele code into the an enthusiastic unsorted .bim not any longer reasons –make-bed to help you segfault.
1 August: –amount logistic regression bugfix. –make-just-bim and you will –make-just-fam flags added. .bim/.fam data files may now become processed in the place of an accompanying .bed less than some situations. –recode-allele today functions properly having A great-transpose setting.
18 July: Text filesets which have each other multi-profile allele codes and you will an enthusiastic unsorted .chart document no more result in the autoconverter so you can crash. –pfilter will be now consistently filter out ‘NA’ records. –dose chromosome code returns bug repaired. –read-freq today tons A1 allele rules while they are lost from the fundamental dataset, in place of erroring in one to problem. –show-tags put in development create.
Very first Cochran-Mantel-Haenszel and you may Breslow-Big date examination added to innovation generate
4 July: Repaired a couple combine pests and this probably caused studies over the past couple examples of an input otherwise matched fileset getting mishandled. ninety alpha/beta builds, and most one or two products was in fact in it, we recommend redoing the combine which have a more recent generate. (Very merges was indeed unaffected, however, best safer than simply disappointed.)
For those who have matched filesets produced with earlier PLINK step one
step one July: Develop has just produced (sorry about that) –data/–gen/–decide to try demand-range parsing insect. –indep not misreports what amount of pruned variants when indeed there are detailed multicollinearity. –fst, –homog, and –oxford-pheno-name flags additional.
twenty seven June: “–recode oxford” no further places completely wrong IDs whenever used in a comparable focus on given that a sample filter. –fill-missing-a2 flag extra. –recode ‘A-transpose’ and you can ‘include-alt’ modifiers added. “–het brief-sample” means added.
20 June: Fixed –mendel zero chromosome password/segfault bug. –merge-record no more means a resource fileset. –fast-epistasis + –parallel bugfix. –indep-pairphase and QFAM sample finished in advancement create.
ten June: –merge-function step 1 today correctly merges forgotten calls that have just one nonmissing name. –r/–r2 chromosome edge handling bugfixes. ‘0X’/’0Y’/’0M’ chromosome codes produced because of the Oxford systems are in fact acknowledged, and then have supported by –output-chr. –vcf-half-name banner put in regulate management of ‘0/.’ VCF GT viewpoints, and you may default conclusion is becoming ‘error’ form to force a conscious ent make.
5 June: –ld-snp-checklist and you will multipass –r/–r2 bugfixes. Nonstandard ‘0/.’ and you can similar VCF GT job opinions are in fact processed once the if they didn’t have the fresh new at the rear of ‘/.’, rather than resulting in a great segfault. (Handling of this situation is generally configurable afterwards, tune in.) –linear/–logistic permutation bugfix (permutation triumph matter array was not initialized to-zero). –serving linear regression placed into development build.
step 3 June: Gzipper no further deletes new production document when becoming requested so you’re able to append (this was causing huge letter-violation data eg –r2 gz to only hold the productivity of your own past citation). Fixed –genome double-missing-telephone call handling insect, and a great –r/–r2 mixed autosomal/nonautosomal data-handling bug. “–dose can be found” and you will –write-serving set in development generate.
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” no further produces room on header range. –prevents is now offering good ‘no-pheno-req’ modifier and that removes the unnecessary phenotype criteria. –annotate put in innovation create.
Recent Comments