Comprehensive assessment of multiple biases in small RNA sequencing reveals significant differences in the performance of widely used methods

Wright, Carrie; Rajpurohit, Anandita; Burke, Emily E.; Williams, Courtney; Collado-Torres, Leonardo; Kimos, Martha; Brandon, Nicholas J.; Cross, Alan J.; Jaffe, Andrew E.; Weinberger, Daniel R.; Shin, Joo Heon

doi:10.1186/s12864-019-5870-3

Table 1 Summary of Results

From: Comprehensive assessment of multiple biases in small RNA sequencing reveals significant differences in the performance of widely used methods

Representative Figure(s)	Assessment	Clontech	Illumina	NEB	NEXTflex	Deduped	Fivepercent
Method information
Figure 1	Designed for adapter ligation mitigation?	Adapter Ligation free	None	PEG	PEG and Randomized adapters	PEG and Randomized adapters	PEG and Randomized adapters
Figure 1	Designed for reverse transcription and PCR bias mitigation?	None	None	None	None	UMI	None
Performance metrics
Similarity
Figure 2	How similar are the quantifications of the same brain sample?	lowest correlations with other methods	Intermediate	Intermediate	Intermediate	Intermediate	Intermediate
Accuracy
Figure 3	Accuracy - similar detection of different sequences	Intermediate	Worst	Worst	Intermediate	Best	Intermediate
Detection Diversity
Figure 4a-b	Mapping rate to miRNA	Worst	Best	Best	Best	Best^a	Best^a
Figure 4c	Detection of miRNA across triplicates at 1000 ng	Intermediate	Intermediate	Intermediate	Intermediate	Best	Worst
Figure 4d	Consistency of miRNA detection at 1000 ng	Worst	Intermediate	Best	Intermediate	Intermediate	Intermediate
Figure 4e	Diversity of detection of isomiRs across triplicates at 1000 ng	Best	Intermediate	Intermediate	Worst	Intermediate	Worst
Figure 4f	Consistency of isomiR detection at 1000 ng	Intermediate	Worst	Best	Intermediate	Intermediate	Intermediate
Figure 4i	False isomiR detection per synthetic sequence	Worst	Intermediate	Intermediate	Best	Best	Best
Figure 4j	False IsomiR expression	Intermediate	Intermediate	Worst	Best	Best	Best
Consistency
Figure 5a	Consistency across batch	Intermediate	Worst	Best	Best	Best	Intermediate
Figure 5b-c	Consistency across triplicates	Intermediate	Worst	Best	Intermediate	Best	Worst
Overall conclusions
		Poor mapping rate, more accurate miRNA quantification, poor isomiR accuracy	Not as accurate, less detection, and less consistency	Good mapping, not as accurate, quite consistent, poor isomiR accuracy	Good mapping, more accurate miRNA quantification, more accurate isomiR quantification	Best performance overall, more accurate miRNA quantification	Deduping improves performance, not just an effect of reducing number of reads

The table depicts the results of the various assessments performed on the 6 small RNA sequencing methods. ^aThe mapping rate of the raw NEXTflex data should reflect the quality for the data used for the Deduped and Fivepercent methods as these are derived from the raw NEXTflex data

Back to article page

ISSN: 1471-2164

Contact us

Submission enquiries: bmcgenomics@biomedcentral.com
General enquiries: ORSupport@springernature.com

BMC Genomics

Contact us