Label-guided seed-chain-extend alignment on annotated De Bruijn graphs

Abstract

The amount of data stored in genomic sequence databases is growing exponentially, far exceeding traditional indexing strategies' processing capabilities. Many recent indexing methods organize sequence data into a sequence graph to succinctly represent large genomic data sets from reference genome and sequencing read set databases. These methods typically use De Bruijn graphs as the graph model or the underlying index model, with auxiliary graph annotation data structures to associate graph nodes with various metadata. Examples of metadata can include a node's source samples (called labels), genomic coordinates, expression levels, etc. An important property of these graphs is that the set of sequences spelled by graph walks is a superset of the set of input sequences. Thus, when aligning to graphs indexing samples derived from low-coverage sequencing sets, sequence information present in many target samples can compensate for missing information resulting from a lack of sequence coverage. Aligning a query against an entire sequence graph (as in traditional sequence-to-graph alignment) using state-of-the-art algorithms can be computationally intractable for graphs constructed from thousands of samples, potentially searching through many non-biological combinations of samples before converging on the best alignment. To address this problem, we propose a novel alignment strategy called multi-label alignment (MLA) and an algorithm implementing this strategy using annotated De Bruijn graphs within the MetaGraph framework, called MetaGraph-MLA. MLA extends current sequence alignment scoring models with additional label change operations for incorporating mixtures of samples into an alignment, penalizing mixtures that are dissimilar in their sequence content. To overcome disconnects in the graph that result from a lack of sequencing coverage, we further extend our graph index to utilize a variable-order De Bruijn graph and introduce node length change as an operation. In this model, traversal between nodes that share a suffix of length < k-1 acts as a proxy for inserting nodes into the graph. MetaGraph-MLA constructs an MLA of a query by chaining single-label alignments using sparse dynamic programming. We evaluate MetaGraph-MLA on simulated data against state-of-the-art sequence-to-graph aligners. We demonstrate increases in alignment lengths for simulated viral Illumina-type (by 6.5%), PacBio CLR-type (by 6.2%), and PacBio CCS-type (by 6.7%) sequencing reads, respectively, and show that the graph walks incorporated into our MLAs originate predominantly from samples of the same strain as the reads' ground-truth samples. We envision MetaGraph-MLA as a step towards establishing sequence graph tools for sequence search against a wide variety of target sequence types.

Indexing All Life’s Known Biological Sequences

“…Then, the respective annotations are retrieved and the aggregated result is returned as output ( Extended Data Figure 1 c ). For increased sensitivity, we developed algorithms for sequence-to-graph alignment 40,46 , which identify the closest matching path in the whole graph ( Extended Data Figures 1 d; Methods ). We also designed a batch query algorithm (schematic in Extended Data Figure 1 e, Methods ), exploiting the presence of k -mers shared between individual queries by forming a fast intermediate query subgraph , that increases throughput up to 100-fold for large repetitive queries (e.g., sets of sequencing reads).…”

“…When label recombination is not desired, we support an alternative approach where queries are aligned to subgraphs of the joint graph induced by single annotation labels (columns of the annotation matrix). We call this approach label-consistent graph alignment (or alignment to columns ) and is implemented by the MetaGraph-LA algorithm 46 . However, instead of aligning to all the subgraphs independently, we perform the alignment with a single search procedure while keeping track of the annotations corresponding to the alignments.…”

“…When mapping to joint graphs, we only considered mapping results that retrieved the ground-truth label of each query read. For all granularities, we mapped the reads via both exact k -mer matching and label-consistent sequence-to-graph alignment using MetaGraph-LA 46 . We measure how well the reads aligned as the percentage of characters in the query that are covered by at least one reported mapping.…”

Prokrustean Graph: A substring index for rapid k-mer size analysis

“…Such exhaustive approaches, sans additional theoretical insights, are unlikely to achieve a significant breakthrough due to inherent computational burden of storing or operating over the representations. Many other tasks addressing similar issues include genome alignment [16, 18, 21] and error correction [1].…”