Positive controls

Validated against the field's reference tools.

A positive control validates an analysis by reproducing a known-correct result. For each Haritica analysis, an independent peer-reviewed reference — or an externally published truth set — is run on the identical public data, and the two outputs are compared. Because the inputs are the same, agreement isolates the analysis engine itself.

= 1.000
log₂FC vs R DESeq2
Jaccard 1.0
enrichment vs clusterProfiler
bit-identical
variants vs NIST GIAB
ρ = 0.92
single-cell vs 10x CellRanger
= 0.966
WGCNA vs R WGCNA
/ 3
Linkage Mapper recovers chr 3
§ The library

The six validations.

Each report pairs a Haritica analysis against an independent reference on the same public inputs. We're publishing the library one study at a time — open a live report to see the full method, figures and caveats.

#
Analysis
Reference
Result
01
R/Bioconductor DESeq2
log₂FC Pearson r = 1.000; DE-call Jaccard 0.987
02
R WGCNA
reciprocal body-weight module; eigengene r = 0.966
03
10x CellRanger + Scanpy
ρ = 0.92 vs CellRanger; per-cell ARI 0.77
04
NIST GIAB HG001; bcftools / SnpEff
bit-identical statistics (Ti/Tv 3.06); validated vs GIAB
05
clusterProfiler + enrichplot
Jaccard = 1.0 across all six gene-set collections
06
Linkage MapperComing soon
MMAPPR2 (Bioconductor)
recovers the published gl13 locus on chr 3 in all three EMS families
§ Method

How each comparison is made.

Each report pairs Haritica against an independent reference run on the identical public inputs, so any difference is attributable to the analysis engine alone. The reference is a peer-reviewed implementation or an externally published truth set — never Haritica's own output. Caveats and the limits of each comparison are stated plainly in every report.

Get started

Reproduce a positive control on day one.

Download Haritica, run one of these analyses on the same public data, then point it at your own. 14-day free trial — no credit card.