As seen in Figure 4a, Bowtie1 can exhibit very low alignment rates for these samples.
Additional file 1: Table S1 shows the alignment rates for all samples.
The median total raw reads for the normal and tumor samples were approximately 11.8 million and 16.3 million separately, and the median alignment rates for the normal and tumor samples were 93.01% and 92.48% separately (Supplementary Table 2 and Supplementary Table 3).
At least 37 million quality-filtered reads for each sample was generated, with an alignment rate of approximately 60% for each library.
Mapping all RNA-seq reads against both transcriptomes (CLC or Velvet/Oases) revealed that the overall mapping success rate, as a measure for assembly quality, was significantly lower for the Velvet/Oases assembly compared to the CLC assembly (see Additional file 1), with an average overall alignment rate of 91.7 % and 73.4 % for the CLC and Velvet/Oases assembly, respectively.
C) Plot showing percent-alignment rates for libraries.
However, descriptive statistics for alignment rate, such as those found in Table 2, should be interpreted with the fact that the test and control groups have significantly different initial LII scores.
By eliminating reads with multiple alignments, the false alignment rate decreases to almost 0 for SOAP2, Bowtie, and BWA, and below 9% for Novoalign (Table 10).
The read alignment rate varied among technologies: 97.4% for TruSeq, 97.7% for NimbleGen, 97.6% Agilent, and 98.95% for Nextera.
For AzoCH3, the alignment rate increased along with the increase of the content of SNPs, showing that the alignment rate, at least for "fast" process, could be enhanced by doping SNPs.
