Algorithmic Bioinformatics

Hannes Hauswedell

Local Aligner for Massive Biological Data

Academic Advisor: Jochen Singer , Knut Reinert
Discipline: Bioinformatik
Degree: Master of Science (M.Sc.)
Degree: Dec 07, 2013
Project:
Status: finished

Abstract

Motivation. Next-generation sequencing technologies produce unprecedented amounts of data, leading to completely new research fields. One of these is metagenomics, the study of large-size DNA samples containing diverse organisms. A key problem in metagenomics is functionally and taxonomically classifying the sequenced DNA, to which end the well known BLAST program is usually used. But BLAST has dramatic resource requirements at metagenomic scales of data, imposing a high financial or technical burden on the researcher. Multiple attempts have been made to overcome these limitations and present a viable alternative to BLAST, but as of yet, they have not gained widespread adoption.

Results. In this work we present Lambda, our own alternative for BLAST in the context of se- quence classification. In our tests Lambda is among the best tools at reproducing BLAST’s results and is faster than all existing solutions at comparable levels of sensitivity. 

Downloads