Quantification of the effects of chimerism on read mapping, differential expression and annotation following short-read de novo assembly.

Background: De novo assembly is often required for analysing short-read RNA sequencing data. An under-characterized aspect of the contigs produced chimerism, extent to which affects mapping, differential expression analysis and annotation. Despite long-read negating this issue, short-reads remain in use through on-going research archived datasets created during last two decades. Consequently, there still a need quantify chimerism its effects. Methods: Effects on mapping were quantified by simulating reads off Drosophila melanogaster cDNA library these related reference sets containing increasing levels chimerism. Next, ten read simulated divided into conditions where, within one, representing 1000 randomly selected transcripts over-represented across replicates. Differential was performed iteratively with set. Finally, an expectation r-squared values describing relationship between alignment transcript lengths matches involving those incrementing created. Similar calculated three graph-based assemblers, relative from input simulated, or sequenced (relative species represented), compared. Results: At 5% 95% sets, 100% 77% mapped, making success poor indicator over-representation, 953 identified analysis; at 10% 936 identified, while it 510. This indicates that despite success, per-transcript counts are unpredictably altered. R-squared obtained assemblers suggest 5-15% chimeric. Conclusions: Although not evident based had significant impact megablast identification. will have consequences past present experiments short-reads.