Skip to content

Validation

bsaseq reproduces published results across two systems that differ in genome size, trait type, and upstream pipeline.

Sorghum ms8 male sterility

Data: Xiao et al. 2025 (Genetics), NCBI BioProject PRJNA1134249 (SRR29793691-94). Default parameters.

Region Max Z Variants
Chr04:61,000,001-63,000,000 12.3 35
Chr04:56,750,001-58,750,000 6.11 131
Chr09 5.69 18
Chr01 4.46 102

The top region contains the causal variant at 61,494,009 (a bHLH transcription factor), matching the published mapping interval (Chr4:58.3-62 Mb). Secondary Chr09/Chr01 peaks are low and consistent with background. The published pipeline used Bowtie2 v2.2.9 + SAMtools v1.11 + bcftools + SnpEff v4.3a.

Rice root-knot nematode resistance

Data: Lahari et al. 2019 (Euphytica), an F2 population segregating for resistance.

Region Max Z Variants
Chr11:28,250,001-29,500,000 3.46 21
Chr09:20,250,001-21,500,000 3.08 1

The Chr11 region falls entirely within the published 23-29 Mb QTL but is ~5x narrower (1.25 Mb vs ~6 Mb). Unlike QTL-seq, bsaseq did not require parental data. The published pipeline used QTL-seq v1.4.4 + BWA-MEM v0.7.12 + FreeBayes v0.9.14.

Performance

Whole-genome runs on a SLURM cluster (Intel Xeon, 16 GB allocation):

Dataset Genome VCF records Passed Windows Runtime Peak memory
Sorghum ms8 ~730 Mb 769,272 140,219 2,613 21.2 s 0.28 GB
Rice nematode ~390 Mb 3,552,893 198,409 1,499 41.3 s 0.25 GB