AdvISER-PYRO: Amplicon Identification using SparsE Representation of PYROsequencing signal.

Abstract

Converting a pyrosequencing signal into a nucleotide sequence appears highly challenging when signal intensities are low (unitary peak heights ) or when complex signals are produced by several target amplicons. In these cases, the pyrosequencing software fails to provide correct nucleotide sequences. Accordingly, the objective was to develop the AdvISER-PYRO algorithm, performing an automated, fast and reliable analysis of pyrosequencing signals that circumvents those limitations. In the current mycobacterial amplicon genotyping application, AdvISER-PYRO performed much better than the pyrosequencing software in the following two situations: when converting Single Amplicon Sample (SAS) signals into a correct single sequence (97.2% versus 56.5%), and when translating Multiple Amplicon Sample (MAS) signals into the correct sequence pair (74.5%). AdvISER-PYRO is implemented in an R package (http://sites.uclouvain.be/md-ctma/index.php/softwares) and can be used in broad range of clinical applications including multiplex pyrosequencing and oncogene re-sequencing in heterogeneous tumor cell samples.