Description of the bug
The results from the nf-core/phaseimpute pipeline using the test_full profile (GLIMPSE2) show the MultiQC report's "Genotype concordance by samples" table to display the Best genotype r-squared and Imputed dosage r-squared as 1.0.
Displaying "1.0" suggests a perfect imputation, so this rounding is misleading:
- MultiQC Report Display: The table shows 1.0 for both r2 metrics.
- Actual Data (AllSamples.txt): The raw GLIMPSE2 concordance output shows the actual values are high but not perfect, e.g., 0.996808 and 0.997202.
Fix: increase the floating-point precision in the MultiQC module for these columns to display at least 4 decimal places.
Command used and terminal output
Relevant files
No response
System information
No response