Meta-analysis: total tallies over variants

[swvanderlaan]

In the summary of the meta-analysis (below) there is a total tally of variants per category. However, this seems to be a bit of (error file reports many more variants that were not found in the reference) and therefore needs double checking.

Example error-file output:

* chrX:86581529:A_G in [ DATA/MODEL1/META/TEMP/meta.all.unique.variants.reorder.split ] is not present in the Variant Annotation File  -- skipping it.
* chrX:86775594:C_T in [ DATA/MODEL1/META/TEMP/meta.all.unique.variants.reorder.split ] is not present in the Variant Annotation File  -- skipping it.
* chrX:86777100:C_A in [ DATA/MODEL1/META/TEMP/meta.all.unique.variants.reorder.split ] is not present in the Variant Annotation File  -- skipping it.
* chrX:86780351:T_C in [ DATA/MODEL1/META/TEMP/meta.all.unique.variants.reorder.split ] is not present in the Variant Annotation File  -- skipping it.
* chrX:86783054:G_T in [ DATA/MODEL1/META/TEMP/meta.all.unique.variants.reorder.split ] is not present in the Variant Annotation File  -- skipping it.
* chrX:86783867:G_A in [ DATA/MODEL1/META/TEMP/meta.all.unique.variants.reorder.split ] is not present in the Variant Annotation File  -- skipping it.
* chrX:86799180:G_T in [ DATA/MODEL1/META/TEMP/meta.all.unique.variants.reorder.split ] is not present in the Variant Annotation File  -- skipping it.
* chrX:86855260:T_C in [ DATA/MODEL1/META/TEMP/meta.all.unique.variants.reorder.split ] is not present in the Variant Annotation File  -- skipping it.
* chrX:86859763:T_C in [ DATA/MODEL1/META/TEMP/meta.all.unique.variants.reorder.split ] is not present in the Variant Annotation File  -- skipping it.
* chrX:86860584:C_A in [ DATA/MODEL1/META/TEMP/meta.all.unique.variants.reorder.split ] is not present in the Variant Annotation File  -- skipping it.

In this example there 4,856 variants skipped. But these are not reported in the summary below as uninformative or something.

Example summary:

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Summarizing this meta-analysis.

* Number of variants in meta-analysis       : 122019.
* Number of variants not in the Reference   : 0.
* Number of uninformative variants skipped  : 0.

          Study name     Allele flips     Sign [beta] flips     Informative variants
          ----------     ------------     -----------------     --------------------
   study1                                   0                     0                   122019
   study2                                   0                     0                   106835

[xEmz]

Read the total lines of the .errors file if that's possible.

[MVPuijk]

Variants that are not in the reference are now counted properly. On the other hand, variants that are uninformative can be counted twice. Like in the example ERROR output below:

* For chr4:69501357:T_TAGAG there is information from one study or none -- so skipping this variant.
* For chr4:69501357:T_TAGAG the effective sample size = 0 -- so skipping this variant.

This results in it counting 2 uninformative variants skipped, despite the fact that these are just 2 errors for the same variant.