Linear Time Construction of Indexable Founder Block Graphs
| Autor: | Mäkinen, Veli, Cazaux, Bastien, Equi, Massimo, Norri, Tuukka, Tomescu, Alexandru I. |
|---|---|
| Přispěvatelé: | Kingsford, Carl, Pisanti, Nadia, Genome-scale Algorithmics research group / Veli Mäkinen, Helsinki Institute for Information Technology, Department of Computer Science, Bioinformatics, Algorithmic Bioinformatics, Carl Kingsford, Nadia Pisanti |
| Rok vydání: | 2020 |
| Předmět: |
0301 basic medicine
founder reconstruction FOS: Computer and information sciences J.3 string matching E.1 F.2.2 education 02 engineering and technology 113 Computer and information sciences compressed data structures 03 medical and health sciences Pangenome indexing Applied computing → Genomics 030104 developmental biology Theory of computation → Sorting and searching Computer Science - Data Structures and Algorithms 0202 electrical engineering electronic engineering information engineering Theory of computation → Dynamic programming multiple sequence alignment 020201 artificial intelligence & image processing Data Structures and Algorithms (cs.DS) Theory of computation → Pattern matching |
| Zdroj: | 20th International Workshop on Algorithms in Bioinformatics (WABI 2020) |
| DOI: | 10.48550/arxiv.2005.09342 |
| Popis: | We introduce a compact pangenome representation based on an optimal segmentation concept that aims to reconstruct founder sequences from a multiple sequence alignment (MSA). Such founder sequences have the feature that each row of the MSA is a recombination of the founders. Several linear time dynamic programming algorithms have been previously devised to optimize segmentations that induce founder blocks that then can be concatenated into a set of founder sequences. All possible concatenation orders can be expressed as a founder block graph. We observe a key property of such graphs: if the node labels (founder segments) do not repeat in the paths of the graph, such graphs can be indexed for efficient string matching. We call such graphs segment repeat-free founder block graphs. We give a linear time algorithm to construct a segment repeat-free founder block graph given an MSA. The algorithm combines techniques from the founder segmentation algorithms (Cazaux et al. SPIRE 2019) and fully-functional bidirectional Burrows-Wheeler index (Belazzougui and Cunial, CPM 2019). We derive a succinct index structure to support queries of arbitrary length in the paths of the graph. Experiments on an MSA of SAR-CoV-2 strains are reported. An MSA of size $410\times 29811$ is compacted in one minute into a segment repeat-free founder block graph of 3900 nodes and 4440 edges. The maximum length and total length of node labels is 12 and 34968, respectively. The index on the graph takes only $3\%$ of the size of the MSA. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|