kmacs: the k -mismatch average common substring approach to alignment-free sequence comparison

Abstract

Motivation: Alignment-based methods for sequence analysis have various limitations if large datasets are to be analysed. Therefore, alignment-free approaches have become popular in recent years. One of the best known alignment-free methods is the average common substring approach that defines a distance measure on sequences based on the average length of longest common words between them. Herein, we generalize this approach by considering longest common substrings with k mismatches. We present a greedy heuristic to approximate the length of such k-mismatch substrings, and we describe kmacs, an efficient implementation of this idea based on generalized enhanced suffix arrays.Results: To evaluate the performance of our approach, we applied it to phylogeny reconstruction using a large number of DNA and protein sequence sets. In most cases, phylogenetic trees calculated with kmacs were more accurate than trees produced with established alignment-free methods that are based on exact word matches. Especially on protein sequences, our method seems to be superior. On simulated protein families, kmacs even outperformed a classical approach to phylogeny reconstruction using multiple alignment and maximum likelihood.Availability and implementation: kmacs is implemented in C++, and the source code is freely available at http://kmacs.gobics.de/Contact: chris.leimeister@stud.uni-goettingen.deSupplementary information: Supplementary data are available at Bioinformatics online.

A wavelet-based feature vector model for DNA clustering

“…Heyne et al (2012) mapped RNA sequence structure information into a graph and clustered the RNAs by means of a graph kernel. Leimeister and Morgenstern (2014) compared DNA sequences based on the longest common substrings with k mismatches. Eisen (1998), Yi et al (2007), and Hatfull et al (2010) clustered genomes by various types of statistical information from DNA microarray data.…”

ALFRED: A Practical Method for Alignment-Free Distance Computation

“…This value is used to estimate pairwise distance of X and Y, in an alignment-free manner, for phylogenetic tree reconstruction (Leimeister and Morgenstern 2014;Horwege et al 2014) Horwege, Lindner, Boden, Hatje, Kollmar, Leimeister, and Morgenstern]. The distance Dist k (X‚ Y), computed as follows, is considered as a good estimate of the evolutionary distance between two species whose corresponding sequences are X and Y, respectively.…”

“…The previous solution to this problem also took near quadratic time (Leimeister and Morgenstern, 2014;Apostolico et al, 2014)Apostolico, Guerra, and Pizzi]. To circumvent this quadratic time barrier, some fast application-specific heuristics (Leimeister and Morgenstern 2014) have already been proposed with near linear time complexity with respect to n and k. Although not as immediate as the above examples, the AAT algorithm is capable of solving another general problem.…”

“…To circumvent this quadratic time barrier, some fast application-specific heuristics (Leimeister and Morgenstern 2014) have already been proposed with near linear time complexity with respect to n and k. Although not as immediate as the above examples, the AAT algorithm is capable of solving another general problem. Let S = fS 1 ‚ S 2 ‚ S 3 ‚ .…”

“…Interestingly, for k = 4 and 5, Robinson-Foulds distance is 0, which means the method is able to reconstruct exactly the same tree topology of the reference tree (Haubold et al, 2009), which is created using alignment. Note that none of the existing alignment-free approaches are able to exactly reconstruct the entire phylogenetic tree (see Leimeister and Morgenstern, 2014), except for the recently published spaced-seed Leimeister, Boden, Horwege, Lindner, and Morgenstern] (but it requires careful parameter tuning). The tree generated using our method with k = 4 is shown in Figure 3.…”