Backcross

Backcross characterizes near-isogenic lines from SNP genotype calls. It compares each line with its recurrent and donor parents, estimates recurrent-parent genome recovery, locates donor segments, checks target regions, and flags genotype quality concerns. The browser app displays graphical genotypes and exports CSV tables and an HTML report. A Node command-line interface uses the same analysis core.

Status: version 0.1.0 (see Releases). Open the app or load the synthetic demo. The demo contains six generated lines and 500 markers; it does not represent a breeding program.

Data and interpretation

The app accepts VCF, HapMap, and wide CSV genotype files, a samples.csv manifest naming one recurrent and one donor parent, and an optional markers.csv genetic map. BrAPI v2.1 allele-matrix loading is also available. Crop-specific chromosome conventions cover soybean, maize, rice, sorghum, wheat, barley, oat, common bean, cotton, cowpea, pea, peanut, and sunflower. The input contract and coding reference specify accepted fields and calls.

Results describe the supplied markers and parent calls. Sparse or uneven marker coverage leaves segment breakpoints uncertain and can change genome-recovery estimates; the report includes the parameters used. Files selected from disk are analyzed in the browser tab. A BrAPI load requests genotype data from the server the user selects. No real genotype dataset is included in this repository.

Run locally

Node.js 22.19 or newer is required.

npm ci
npm run dev

Open the local address printed by Vite. For batch output, the synthetic fixture provides a working CLI example:

node src/cli.ts summarize --genotypes tests/fixtures/synthetic/genotypes.vcf --samples tests/fixtures/synthetic/samples.csv --markers tests/fixtures/synthetic/markers.csv --out summary.csv

Every GitHub Release also carries backcross-cli-<version>.mjs, a single file that needs only Node 22.19: node backcross-cli-<version>.mjs summarize ... with the same options.

The segments, targets and qc commands produce separate CSV files; compare and discordant write the pairwise comparison and discordant-marker CSVs for one pair of sample ids given as --a and --b (--mode informative, the default, or all). Every command takes --crop ID (one of the thirteen chromosome schemes, default soybean) and --profile ID|FILE (a token profile for HapMap and wide CSV, or a JSON file of the same shape); run node src/cli.ts for the full option list. npm run build creates a static site in dist/.

Verification and documentation

npm test, npm run typecheck, npm run lint, and npm run test:browser check the analysis and interface. The synthetic fixture has independently generated expected values; the plan records measured performance and remaining verification limits. The user guide, the docs index, data formats and exports, design decisions, and the archived README provide detail.

Cite

DOI for all versions: 10.5281/zenodo.23051345 (version 0.1.0: 10.5281/zenodo.23051346).

Cite the version you used: CITATION.cff carries the metadata GitHub's "Cite this repository" button and Zenodo read, and every CSV export and HTML report record the release and commit that produced them (tool_version, tool_commit; docs/data-formats.md). The software is available under the MIT license.

Sibling tool

Backcross characterises finished near-isogenic lines; progeny-selector ranks progeny during the programme, generation by generation. Both read input data contract 1.12.0 (contract/data-contract.md), so genotype, samples.csv and markers.csv files move between them unchanged.

Contributions follow CONTRIBUTING.md and the code of conduct; vulnerabilities go through SECURITY.md.