vampire.anno.tl.sample_msa

Contents

vampire.anno.tl.sample_msa#

vampire.anno.tl.sample_msa(adata, *, match_score=2, mismatch_penalty=-3, gap_open_penalty=-5, gap_extend_penalty=-1, refine=True, max_refine_iter=3, store_key='aligned')[source]#

Multiple sequence alignment of motif arrays across samples.

Uses a guide-tree progressive alignment strategy: build a UPGMA tree from pairwise distances, then merge profiles bottom-up. Optionally refines the MSA via iterative consensus realignment.

Duplicate sequences are automatically deduplicated before alignment and mapped back afterwards.

Parameters:

  • adata (AnnData) – Annotated data with motif_array and orientation_array in uns, and motif_distance in varp.

  • match_score (int) – Reward coefficient for matching motifs. The substitution score is (avg_len - distance) * match_score + distance * mismatch_penalty, where avg_len is the average motif length of the two motifs.

  • mismatch_penalty (int) – Penalty coefficient for mismatched motifs.

  • gap_open_penalty (int) – Penalty for opening a gap.

  • gap_extend_penalty (int) – Penalty for extending a gap.

  • refine (bool) – Whether to perform iterative consensus-based refinement after the initial progressive alignment.

  • max_refine_iter (int) – Maximum number of refinement iterations. Ignored when refine=False.

  • store_key (str) – Key prefix for storing results in adata.uns. Stores {store_key}_motif_array, {store_key}_orientation_array, {store_key}_consensus, and {store_key}_consensus_orientation.

Returns:

The updated AnnData with alignment results in uns.

Return type:

AnnData

Examples

>>> import vampire as vp
>>> adata = vp.datasets.wdr7_hprc()
>>> vp.anno.tl.sample_msa(adata)
Contents