The Closest BLAST Hit Is Often Not the Nearest Neighbor

By L. Koski, G. B. Golding. Journal of Molecular Evolution, 2001

507 citations recorded by the source corpus.

Subfields: Molecular Biology, Plant Science

View Paper: https://www.semanticscholar.org/paper/a1f3306999f01ed443ba95526716824969767a49

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Molecular Biology

In Plant Science

Citations in the corpus

Gene transfer into the fungi (2011, background): “Of these, 24 genes demonstrated phylogenies consistent with vertical transmission, while 15 of the sub-sample of vertically transmitted genes had previously been classified as HGTs based upon composition data (Koski et al., 2001; Lawrence and Ochman, 1998).”
Testing the reliability of genetic methods of species identification via simulation. (2008, background): “Little and Stevenson (2007) found that BLAST can give accurate identification at the genus, but not the species, level in gymnosperms, perhaps because top hits are often not closest phylogenetic relatives (Koski and Golding, 2001).”
SyMAP v3.4: a turnkey synteny system with application to plant genomes (2011, background): “Traditional reciprocal-best-hit filtering corresponds to N=1, but a value of a least 2 aids in detecting duplications [also, N=1 rejects many true orthologs (35)].”
Genomics of bacteria and archaea: the emerging dynamic view of the prokaryotic world (2008, unknown): “Furthermore, it has been shown that phylogenetic analysis often does not support the conclusions on evolutionary relationships drawn from sequence similarity analysis suggesting that some of the conclusions drawn from BLAST-based comparisons could be misleading (183).”
Choosing BLAST options for better detection of orthologs as reciprocal best hits (2008, methodology): “However, the comparison of BLAST hits with evolutionary neighbors was based on a default BLAST run, with evolutionary neighbors detected using phylogenetic trees (Koski and Golding, 2001), not just distances calculated from amino acid substitutions (as in Wall et al., 2003).”
Assessing Performance of Orthology Detection Strategies Applied to Eukaryotic Genomes (2007, background): “The relatively high FP rate observed for SBH might be expected, as one-way best hits are frequently not the nearest neighbor [32].”
Phylogenomics and the reconstruction of the tree of life (2005, methodology): “This assessment is primarily based on sequence-similarity searches that can be misleadin”
A comparison of algorithms for the identification of specimens using DNA barcodes: examples from gymnosperms (2007, background): “Some query sequence misidentifications can be ascribed to an incomplete reference database (Koski and Golding, 2001) or the parameters used to construct pair-wise alignments (Anderson and Brass, 1998; Woodwark et al.”
GoFDR: A sequence alignment based method for predicting protein functions. (2016, background): “In the easiest case, if the sequence identity of a known sequence is above 60% to the query sequence, then the function of the known sequence can be transferred to the query sequence with above 90% accuracy [33,34].”
Clustering metagenomic sequences with interpolated Markov models (2010, background): “However, sequence alignment can only classify reads from organisms with a close evolutionary relative that has already been sequenced [23].”
The unholy trinity: taxonomy, species delimitation and DNA barcoding (2005, background): “A second shortcoming is that similarity scores often do not give the nearest neighbour as the closest relative ( Koski & Golding 2001 ).”
CLANS: a Java application for visualizing protein families based on pairwise similarity (2004, background): “However, top-scoring BLAST matches do not necessarily identify the closest homologs (Koski and Golding, 2001).”
DNA barcoding using chitons (genus Mopalia) (2007, background): “While rapid and popular, approaches such as blast have some primary drawbacks: the occurrence of high rates of ‘false-positives’, and sampling-dependent accuracy is prevalent ( Koski & Golding 2001 ).”
Horizontal gene transfer, genome innovation and evolution (2005, methodology): “The use of phylogenetic reconstruction promises more reliable detection of HGT events than simple database searche”
Metagenomics: DNA sequencing of environmental samples (2005, methodology): “The simplest method of taxonomic assignment, best BLAST hit, should be used with caution: it is only reliable when close relatives are available for comparison, and is essentially useless when no relatives have been fully sequence”
A Bioinformatician's Guide to Metagenomics (2008, background): “The closest BLAST hit is not necessarily the nearest phylogenetic neighbor (72), and therefore, classifying by BLAST hits can be misleading, particularly if only distantly related homologs are available in the reference database.”