deletions | additions       diff --git a/how the mapping works.tex b/how the mapping works.tex  index 3d7aab3..82c2661 100644  --- a/how the mapping works.tex  +++ b/how the mapping works.tex 
...
\subsection{How WASP mapping works}  In the WASP mapping pipeline, the user first maps reads to the genome using a mapper of their choice that outputs outputs BAM or SAM format. For example, ChIP-seq reads can be mapped by BWA \cite{Li_2009} or Bowtie \cite{dx.doi.org/doi:10.1186/gb-2009-10-3-r25}, \cite{10.1186/gb-2009-10-3-r25},  and RNA-seq reads can be mapped using tophat \cite{Trapnell_2009}. WASP then identifies all mapped reads that overlap with known polymorphisms. For each read, all possible allelic combinations that differ from the original read are generated and these reads are re-mapped to the genome. For example, when a read overlaps two bi-allelic SNPs four allelic combinations are possible, three of which differ from the original read. If any of the generated reads fails to map uniquely to the same location as the original read, the original read is discarded. This simple method has the advantage that it works with almost any existing mapping pipeline and handles reads with sequencing errors, which can be a major source of biased mapping \cite{Degner_2009}.