SeqAn an efficient, generic C++ library for sequence analysis

Abstract

Background: The use of novel algorithmic techniques is pivotal to many important problems in life science. For example the sequencing of the human genome 1 would not have been possible without advanced assembly algorithms. However, owing to the high speed of technological progress and the urgent need for bioinformatics tools, there is a widening gap between state-of-the-art algorithmic techniques and the actual algorithmic components of tools that are in widespread use.

Results: To remedy this trend we propose the use of SeqAn, a library of efficient data types and algorithms for sequence analysis in computational biology. SeqAn comprises implementations of existing, practical state-of-the-art algorithmic components to provide a sound basis for algorithm testing and development. In this paper we describe the design and content of SeqAn and demonstrate its use by giving two examples. In the first example we show an application of SeqAn as an experimental platform by comparing different exact string matching algorithms. The second example is a simple version of the well-known MUMmer tool rewritten in SeqAn. Results indicate that our implementation is very efficient and versatile to use.

Conclusion: We anticipate that SeqAn greatly simplifies the rapid development of new bioinformatics tools by providing a collection of readily usable, well-designed algorithmic components which are fundamental for the field of sequence analysis. This leverages not only the implementation of new algorithms, but also enables a sound analysis and comparison of existing algorithms.

Figure 1

Genome comparison tools and their algorithmic components.

Figure 2

SeqAn Contents Overview.

Figure 3

Runtimes of String Matching Algorithms. We compared three exact string matching algorithms from SeqAn with the member function basic_string::find of the standard library, as it was implemented for Microsoft Visual C++. The left figure shows the runtimes (in ms) for searching a DNA sequence (human chromosome 21), the right figure for searching a proteine database. The search pattern was taken randomly from the sequence. The figures show the average time needed to find all occurrences of patterns of a given length.