APPLES: Fast Distance-Based Phylogenetic Placement

摘要

Methods for inferring phylogenetic trees from very large datasets exist, yet, large-scale tree reconstructions still require significant resources. New species are continually being sequenced, and as a result, even large trees can become outdated. Reconstructing the tree de novo each time new sequences become available is not practical. An alternative approach is phylogenetic placement where new sequence(s) are simply added to an existing backbone tree. Phylogenetic placement has applications other than updating trees, including sample identification, where the goal is to detect the identity of given query sequences of unknown origins. This problem arises [3] in the study of mixed environmental samples that make up much of the microbiome literature. Sample identification is also the essence of barcoding and meta-barcoding, methods used often in biodiversity studies.