Bug in plink2, getting out of memory in plink2, but plink 1.07 can process similar command

[DK]

Success with below command in plink 1.07 (this was just a test for 1 snp, I am aware that sex is missing for this analysis, but it demonstrates the point):

options in effect:

        --noweb

        --file FV

        --pheno  ./pheno_plink.txt

        --logistic

        --sex

        --out analysis.fv.param

        --covar ./covar_plink.txt

        --covar-number 1-6

** For gPLINK compatibility, do not use '.' in --out **

1 (of 1) markers to be included from [ FV.map ]

Warning, found 152249 individuals with ambiguous sex codes

These individuals will be set to missing ( or use --allow-no-sex )

Writing list of these individuals to [ lysis.fv.param.nosex ]

152249 individuals read from [ FV.ped ]

0 individuals with nonmissing phenotypes

Assuming a disease phenotype (1=unaff, 2=aff, 0=miss)

Missing phenotype value is also -9

0 cases, 0 controls and 152249 missing

0 males, 0 females, and 152249 of unspecified sex

Reading alternate phenotype from ./pheno_plink.txt ]

152095 individuals with non-missing alternate phenotype

Assuming a disease phenotype (1=unaff, 2=aff, 0=miss)

Missing phenotype value is also -9

4060 cases, 148035 controls and 154 missing

Reading 6 covariates from [ ./covar_plink.txt ] with nonmissing values for 152248 individuals

Selected subset of 6 from 6 covariates

For these, nonmissing covariate values for 152248 individuals

Before frequency and genotyping pruning, there are 1 SNPs

152249 founders and 0 non-founders found

Total genotyping rate in remaining individuals is 1

0 SNPs failed missingness test ( GENO > 1 )

0 SNPs failed frequency test ( MAF < 0 )

After frequency and genotyping pruning, there are 1 SNPs

After filtering, 4060 cases, 148034 controls and 155 missing

After filtering, 0 males, 0 females, and 152249 of unspecified sex

Converting data to Individual-major format

Writing logistic model association results to [.assoc.logistic ]

Analysis finished: Fri Jan 29 11:22:53 2016

-- Same code, in plink2, on  a test server with MASSIVE amounts of RAM, only difference is that i used the --parameters flag for covariate selection instead of covar-number.  I tried --covar-number flag in plink2 as an alternative, the regression results are extremely error prone (i have a positive control run in R, and the results in plink2 are wildly disparate). Also, i have a --pheno flag with a pheno file, but plink2 doesnt recognize it.  I also tried to replace the --linear sex with just linear, same results

PLINK v1.90b1g 64-bit (10 Jun 2014)        https://www.cog-genomics.org/plink2

(C) 2005-2014 Shaun Purcell, Christopher Chang   GNU General Public License v3

Logging to lysis.fv.param.log.

Note: --noweb has no effect since no web check is implemented yet.

4004741 MB RAM detected; reserving 2002370 MB for main workspace.

.ped scan complete (for binary autoconversion).

Performing single-pass .bed write (1 variant, 152249 people).

--file: temporary.bed +-temporary.bim +-temporary.fam written.

1 variant loaded from .bim file.

152249 people (0 males, 0 females, 152249 ambiguous) loaded from .fam.

Ambiguous sex IDs written to param.nosex .

152095 phenotype values present after --pheno.

Warning: Ignoring phenotypes of missing-sex samples.  If you don't want those

phenotypes to be ignored, use the --allow-no-sex flag.

Using 1 thread (no multithreaded calculations invoked).

--covar: 6 covariates loaded.

1 person had missing value(s).

Calculating allele frequencies... done.

1 variant and 152249 people pass filters and QC.

Note: No phenotypes present.

Warning: Ignoring --linear/--logistic 'sex' modifier since sex is invariant.

Warning: Skipping --linear since # variables >= # samples.

Error: Out of memory.  Try the --memory and/or --parallel flags.

[Christopher Chang]

That build is more than 19 months old; there is a long list of bugs which have been fixed since then.  Can you check if this works with the latest stable build (b3.30)?