Feature Request: Add Confidence Intervals for Motif Counts (MC)

[nlfortier]

The current tandem repeat genotyping tool outputs motif counts (MC) in VCF files but does not provide confidence intervals for these estimates. This would be a valuable addition since:

• Confidence intervals help users assess call reliability
• This feature exists in other STR callers (e.g., ExpansionHunter's REPCI field)
• It would improve consistency with community standards

Example from ExpansionHunter

##FORMAT=<ID=REPCI,Number=1,Type=String,Description="Confidence interval for REPCN">
##FORMAT=<ID=REPCN,Number=1,Type=String,Description="Number of repeat units spanned by the allele">

In the VCF records:

REPCN:REPCI 2/10:2-4/9-12

Suggested Implementation:

• Add a new FORMAT field (e.g., MCCI) for motif count confidence intervals
• Format as two integers (min,max) per genotype like ExpansionHunter's REPCI
• Document the confidence level used (e.g., 95% CI)

Benefits:

• Enables better quality filtering
• Facilitates comparison with other tools
• Helps clinical/lab users make informed decisions

Would this be feasible to implement in a future release?

[pbsena]

Hello,

Thank you for the suggestion. Just to confirm, the ALLR field shows the range of read lenghts per allele (i.e., the min and max length for reads assigned to each allele). We renamed from ALCI to be more in line with the actual definition of the values reported. Do you need the 95% CI in addition to the ranges?

Regardless I agree that this is a good addition to differentiate the presence of outliers from an actually mosaic range in simple summary statistics.