The accurate identification of genetic variants in single-cell genomes could enable a variety of clinical applications. Currently, accurately detecting such variants remains challenging due to the limitations of DNA sequencing methods.

Researchers report a method for improving the accuracy of single-cell genome sequencing and haplotyping called single-stranded sequencing using microfluidic reactors (SISSOR). The method uses a microfluidic processor that consists of modules for single-cell capture, cell lysis and strand separation, partitioning, and amplification.

Once single-cell double-stranded chromosomal DNA molecules are separated, the DNA fragments are randomly distributed and partitioned into 24 identical nanoliter compartments for amplification using multiple displacement amplification techniques. The amplified products are then retrieved from each compartment, converted to barcoded sequencing libraries, and sequenced with Illumina short-read sequencing-by-synthesis methods.

The authors demonstrated SISSOR using three single cells from the human PGP1 fibroblast cell line. Overall, the error rate of SISSOR sequencing was below 10−8, yielding 4 possible errors in 351 Mb.

Prior to haplotype assembly, the average DNA fragment length was approximately 500 kb, 5–10-fold longer than lengths achieved using dilution methods. According to the authors, the method might improve the accuracy of single-cell genome sequencing and provide longer haplotype lengths than current approaches.

http://www.pnas.org/content/early/2017/10/23/1707609114