2021. 02NGS
Performance comparison of four types oftarget enrichment baits for exome DNA sequencing
Juan Zhou, Mancang Zhang, Xiaoqi Li, Zhuo Wang, Dun Pan and Yongyong Shi
Full Text
ABSTRACT
Background: Next-generation sequencing technology is developing rapidly and target capture sequencing hasbecome an important technique. Several different platforms for library preparation and target capture with differentbait types respectively are commercially available. Here we compare the performance of the four platforms withdifferent bait types to find out their advantages and limitations. The purpose of this study is to help investigatorsand clinicians select the appropriate platform for their particular application and lay the foundation for thedevelopment of a better target capture platform for next-generation sequencing.
Results: We formulate capture efficiency as a novel parameter that can be used to better evaluations of specificityand coverage depth among the different capture platforms. Target coverage, capture efficiency, GC bias, ATDropout, sensitivity in single nucleotide polymorphisms, small insertions and deletions detection, and the feature ofeach platform were compared for low input samples. In general, all platforms perform well and small differencesamong them are revealed. In our results, RNA baits have stronger binding power than DNA baits, and with ultradeep sequencing, double stranded RNA baits perform better than single stranded RNA baits in all aspects. DNAbaits got better performance in the region with high GC content and RNA baits got lower AT dropout suggestingthat the binding power is different between DNA and RNA baits to genome regions with different characteristics.
Conclusions: The platforms with double stranded RNA baits have the most balanced capture performance. Ourresults show the key differences in performance among the four updated platforms with four different bait types.The better performance of double stranded RNA bait with ultra deep sequencing suggests that it may improve thesensitivity of ultra low frequent mutation detection. In addition, we further propose that the mixed baits of doublestranded RNA and single stranded DNA may improve target capture performance