Error: --pmerge[-list] is under development.

[Peiyuan Zhu]

Hi Christopher, I have tried using a txt file to contain the file names, which I suppose is what is meant by the documentation. However, I still get the error saying this function is under development, which I suppose this won't happen because I'm already using the latest version. 

[zhupy@blg4110 zhupy]$ cat ukb_c.txt

ukb_c1

ukb_c2

ukb_c3

ukb_c4

ukb_c5

ukb_c6

ukb_c7

ukb_c8

ukb_c9

ukb_c10

ukb_c11

ukb_c12

ukb_c13

ukb_c14

ukb_c15

ukb_c16

ukb_c17

ukb_c18

ukb_c19

ukb_c20

ukb_c21

ukb_c22

ukb_cX

ukb_cXY

[zhupy@blg4110 zhupy]$ ./plink2 --pmerge-list ukb_c.txt --pmerge-list-dir . --out ukb_imp_genom --multiallelics-already-joined

PLINK v2.00a3LM 64-bit Intel (28 Mar 2021)     www.cog-genomics.org/plink/2.0/

(C) 2005-2021 Shaun Purcell, Christopher Chang   GNU General Public License v3

Logging to ukb_imp_genom.log.

Options in effect:

  --multiallelics-already-joined

  --out ukb_imp_genom

  --pmerge-list ukb_c.txt

  --pmerge-list-dir .

Start time: Mon Apr  5 02:39:07 2021

192029 MiB RAM detected; reserving 96014 MiB for main workspace.

Using up to 40 threads (change this with --threads).

--pmerge-list: 24 filesets specified.

--pmerge-list: 487409 samples present.

--pmerge-list: Merged .psam written to ukb_imp_genom.psam .

--pmerge-list: 24 .pvar files scanned.

Error: --pmerge[-list] is under development.

End time: Mon Apr  5 02:39:40 2021

[Christopher Chang]

0. Why on earth is "--multiallelics-already-joined" in this command line?

1. The very first line of the --pmerge[-list] online documentation is "(Only handles concatenation-like jobs for now.)" in red.  Take a closer look at the last two filesets in your list.

[Peiyuan Zhu]

Does concatenation-like jobs means the same sample ID should present in all files? The --multiallelics-already-joined flag comes from an error message below. 

I tried removing the last two filesets from X and XY chromosome so I have 24-2=22 filesets overall but still doesn't work. 

[zhupy@blg4110 zhupy]$ ./plink2 --pmerge-list ukb_c.txt --pmerge-list-dir . --out ukb_c

PLINK v2.00a3LM 64-bit Intel (28 Mar 2021)     www.cog-genomics.org/plink/2.0/

(C) 2005-2021 Shaun Purcell, Christopher Chang   GNU General Public License v3

Logging to ukb_c.log.

Options in effect:

  --out ukb_c

  --pmerge-list ukb_c.txt

  --pmerge-list-dir .

Start time: Mon Apr  5 13:46:01 2021

192029 MiB RAM detected; reserving 96014 MiB for main workspace.

Using up to 40 threads (change this with --threads).

--pmerge-list: 22 filesets specified.

--pmerge-list: 487409 samples present.

--pmerge-list: Merged .psam written to ukb_c.psam .

Error: The biallelic variants with ID 'rs151120166' at position 1:715142 in

./ukb_c1.pvar appear to be the components of a 'split' multiallelic variant; if

so, it must be 'joined' (with e.g. "bcftools norm -m") before a correct merge

can occur. If you are SURE that your data does not contain any same-position

same-ID variant groups that should be joined, you can suppress this error with

--multiallelics-already-joined.

End time: Mon Apr  5 13:46:29 2021

[zhupy@blg4110 zhupy]$ ./plink2 --pmerge-list ukb_c.txt --pmerge-list-dir . --out ukb_c --multiallelics-already-joined

PLINK v2.00a3LM 64-bit Intel (28 Mar 2021)     www.cog-genomics.org/plink/2.0/

(C) 2005-2021 Shaun Purcell, Christopher Chang   GNU General Public License v3

Logging to ukb_c.log.

Options in effect:

  --multiallelics-already-joined

  --out ukb_c

  --pmerge-list ukb_c.txt

  --pmerge-list-dir .

Start time: Mon Apr  5 13:46:42 2021

192029 MiB RAM detected; reserving 96014 MiB for main workspace.

Using up to 40 threads (change this with --threads).

--pmerge-list: 22 filesets specified.

--pmerge-list: 487409 samples present.

--pmerge-list: Merged .psam written to ukb_c.psam .

--pmerge-list: 22 .pvar files scanned.

Concatenation job detected.

Concatenating... 0/92775306 variants complete.Error: --pmerge[-list] multiallelic-variant dosage support is under development.

End time: Mon Apr  5 13:47:17 2021

[Peiyuan Zhu]

The ids are duplicated for multialletic variants. 

> ukb_c1 %>% filter(ID=="rs151120166")

# A tibble: 2 x 5

  `#CHROM`    POS ID          REF   ALT  

     <dbl>  <dbl> <chr>       <chr> <chr>

1        1 715142 rs151120166 G     A    

2        1 715142 rs151120166 G     T    

[Christopher Chang]

1. Sample IDs actually aren't required to match.  Instead, the current requirement is that the filesets don't cover overlapping position ranges.  If ukb_cX covers the middle of chrX, and ukb_cXY covers the beginning and end of chrX, that doesn't work for now.  If that is the case, you might be able to address it by splitting ukb_cXY into e.g. "ukb_PAR1" and "ukb_PAR2" parts, and changing the bottom of your list to

  ukb_PAR1

  ukb_cX

  ukb_PAR2

2. However, the real issue is that your data *does* contain split variants.  You are LYING to plink2 with --multiallelics-already-joined, and are LUCKY that your run with that flag happened to fail for another reason; otherwise, you would have destroyed every affected variant.

Eventually, the right thing to do will be (i) join your variants with e.g. "bcftools norm -m +" (there will be an equivalent plink2 function fairly soon), and then (ii) use either "bcftools concat" or --pmerge-list on the results.  However, this requires "multiallelic dosage" functionality that has not been built out yet.

For now, the simplest solution is to use --set-all-var-ids to assign unique IDs to each part of a split variant.

[Peiyuan Zhu]

1. What is the best way of doing the splitting of Chromosome XY?

2. I think this should be done outside of the pmerge command. So I will be doing this when I am converting bgen to pgen. I always have this problem of not knowing which two arguments can be combined into one run. For example, in this case it wouldn't help if I do "./plink2 --set-all-var-ids @:#[b19]\$r,\$a --pmerge-list ukb_c.txt --pmerge-list-dir . --out ukb_c --multiallelics-already-joined". But sometimes arguments can be combined like I can do --max-maf with glm or do  --max-maf to prune and then use the pruned file in glm. I'm just trying to understand how such distinction can be made

[Christopher Chang]

Why is --multiallelics-already-joined still in that command line?!

I'll wait until next week to respond to the other questions.  You need to spend more time understanding the answers I've already given.

[Peiyuan Zhu]

I was assuming  --set-all-var-ids @:#[b19]\$r,\$a is done before  --multiallelics-already-joined because I put that argument first. This is part of the question that is I'm not sure when can we combine arguments into one run. 

[Christopher Chang]

Please reread what I have written about --multiallelics-already-joined.