TandemTwister: scalable genotyping and advanced visualization of tandem repeats.

Copy number variation of tandem repeat units, including pathogenic expansions, genotyped from long-read data and assembled genomes.

Tandem repeats are implicated in complex traits and genetic disorders, including neurodegenerative and developmental diseases, and the growing number of described loci calls for fast, scalable genotyping tools. The authors present TandemTwister, a parallelised C++ algorithm for tandem repeat copy number genotyping, together with an interactive visualisation tool showing exact motif occurrences, counts and population information across haplotypes. Benchmarked in an Ashkenazim trio on 1.2 million annotated tandem repeat regions and across sequencing technologies, it was the fastest and most accurate of the tools compared: on PacBio HiFi data it ran in 15 minutes on 32 CPU cores with 99.4% recall, 98.0% Mendelian consistency and 94% sequence accuracy. The authors also demonstrate super-population clustering and examine the inheritance of tandem repeats and haplotype blocks in three trios. Applied to a cohort of 31 individuals with neurodegenerative and developmental disorders, the tool distinguished normal from pathogenic copy numbers.

Synthesis written by Geno'X. For the full original abstract, please refer to the source publication.

The reported performance is mainly about genotyping accuracy and speed, which does matter when 1.2 million loci are processed routinely, but the clinical demonstration is thin: 31 patients, with no detail on the loci involved or on missed cases. The most useful element for interpretation is the display of the exact motif, since motif composition or interruption often changes what an expansion means. To bring it into a diagnostic pipeline, one would need a targeted evaluation on known pathogenic loci with expansions confirmed by a reference method, plus an estimate of false negatives.

Analysis by Dr Thibaut Benquey