Function reference¶

The complete, always-current documentation for every function and argument is available from R via ?functionName and help(package = "rabhit"). This page summarizes the exported API.

Haplotype inference¶

`createFullHaplotype()`¶

Anchor-based V-D-J haplotype inference — the main entry point.

createFullHaplotype(clip_db, toHap_col = c("v_call", "d_call"),
                    hapBy_col = "j_call", hapBy = "IGHJ6", toHap_GERM = NULL,
                    relative_freq_priors = TRUE, kThreshDel = 3, rmPseudo = TRUE,
                    deleted_genes = c(), nonReliable_Vgenes = c(),
                    min_minor_fraction = 0.3, single_gene = TRUE,
                    chain = c("IGH", "IGK", "IGL", "TRB"))

Key arguments: toHap_col the segment columns to haplotype; hapBy_col/hapBy the anchor column and gene; toHap_GERM the germline (defaults to the bundled set for chain); kThreshDel the Bayes-factor threshold for calling deletions; min_minor_fraction the minimum minor-allele fraction for a usable anchor. Returns a haplotype table with, per gene, the allele assigned to each anchor chromosome, the candidate alleles, read counts, and the Bayes factor lK.

`createHaplotypeTable()`¶

Lower-level routine that haplotypes a single gene against the anchor counts using the Dirichlet-multinomial Bayesian model. Called internally by createFullHaplotype() per gene.

`readHaplotypeDb()`¶

Read a previously saved haplotype table back into R.

`convertToASC()`¶

Prepare data and germline for haplotype inference using PIgLET Allele Similarity Clusters (ASC), which collapse near-identical/duplicated IGHV alleles and align with OGRDB reference sets. Returns ASC-named clip_db, ASC germline, a genes_order, and the allele_cluster_table. Requires the optional piglet package. See Allele clusters (ASC/OGRDB).

Gene usage and deletions¶

`geneUsage()`¶

geneUsage(clip_db, chain = c("IGH", "IGK", "IGL", "TRB"),
          genes_order = NULL, rmPseudo = TRUE)

Per-gene usage frequencies, used as input to the deletion tests.

`deletionsByBinom()`¶

Double-chromosome deletion detection across a population via a binomial test on gene usage.

`deletionsByVpooled()`¶

Single-chromosome deletion detection for V genes using a pooled approach.

`nonReliableVGenes()`¶

Identify V genes that are unreliable (frequently appearing in low-confidence multi-assignments), e.g. in partial-coverage data. The result is passed to deletionsByBinom() and createFullHaplotype().

Visualization¶

Function	Output
`plotHaplotype()`	Per-subject haplotype map (static, or interactive with `html_output = TRUE`)
`hapHeatmap()`	Multi-sample allele-assignment heatmap; returns `list(p, width, height)`
`hapDendo()`	Dendrogram + heatmap clustering samples by haplotype similarity (needs `ggdendro`)
`deletionHeatmap()`	Single-chromosome deletion heatmap across subjects (static or interactive)
`plotDeletionsByBinom()`	Double-chromosome deletions across the population
`plotDeletionsByVpooled()`	V/D/J single-chromosome deletions

Common arguments: chain, genes_order (override gene ordering), removeIGH/removeIGH-style label trimming, lk_cutoff/kThreshDel (confidence thresholds), and html_output for the interactive variants (requires plotly + htmlwidgets).

Bundled data¶

samples_db (example IGH repertoire), samplesHaplotype (example result), the germline references HVGERM/HDGERM/HJGERM/KVGERM/KJGERM/LVGERM/LJGERM, and GENE.loc (chromosomal gene ordering). See Input format.

Function reference¶

Haplotype inference¶

createFullHaplotype()¶

createHaplotypeTable()¶

readHaplotypeDb()¶

convertToASC()¶