Sample Overlap equal to 1

[cburik]

Hi,

I'm using the sample overlap function while doing a meta-analysis on two files and I'm getting an odd result. The estimated sample overlap is equal to 1, when I'm sure it shouldn't be.

best,

Casper

Below is a snip of the log

[...]
## Meta-analysis will be based on sample sizes, p-values and direction of effect ...
## Tracking of extreme allele frequencies enabled
## Averaging of allele frequencies enabled
## Genomic control correction of input statistics disabled
## LENIENT MODE: Input lines can include extra columns, which will be ignored
## Strand flipping according to strand column disabled
## Tracking of chromosomes and positions is enabled
## Correction for samples overlap enabled
## Set column separator to TAB ...
## Set marker header to rsID ...
## Set allele headers to EFFECT_ALLELE and OTHER_ALLELE ...
## Set effect header to BETA ...
## Set p-value header to PVAL ...
## Set standard error header to SE ...
## Set frequency header to EAF ...
## Set weight header to N ...
## Set chromosome header to chr ...
## Set position header to position ...
###########################################################################
## Processing file [...]
## Processed 6144122 markers ...

###########################################################################
## Processing file [...]
## Samples overlap for 2808863 marker(s) with 20000<=WEIGHT<21000: 1.0000000000
## Processed 6024671 markers ...
[...]

[welchr]

Any chance we could use your data for debugging purposes?

[cburik]

Sure, where can I send it?

[welchr]

We could maybe setup an SFTP or a Dropbox file request (as long as the data is less than a few TB.) Shoot me an email at [email protected] and we'll figure something out.