Complete Genomics announced that it has developed a new sequencing technology that improves the accuracy of whole genome sequencing with reduced amount of DNA. The new Complete Genomics’ technology known as Long Fragment Read (LFR) technology allows Complete Genomics to solve phasing genomes, one of the important bottlenecks of existing short read sequencing.
Human genomes are diploid; each chromosome is present in two copies as homologous pair. And fifty percent of these homologous chromosome is from each parent. The sequencing reads from short read technology do not allow us to tell from which parent a read came from, i.e. does not allow us to fully phase the genomes.
Although recent sequencing technology advances have enabled lower cost sequencing, till now only four human personal genomes have been fully reconstructed as diploid (J. Craig Venter, a Gujarati Indian from HapMap project, and two Europeans). In a paper published in Nature journal, Complete Genomics presents LFR technology and shows that it produces highly accurate, fully phased genomes with small DNA sample. The new technology only needs 10 to 20 cells (only 100 picograms of DNA), compared to current technologies needing a few nanograms of DNA and has error rate 1 in 10 million base pairs.
Complete Genomics hopes that the new highly accurate LFR technology needing reduced amount of DNA should help accelerate the use of whole genome sequencing at clinical settings to treat patients. Complete Genomics has already secured two patents on LFR technology, and additional patent applications are pending at the US patent office. In the press release announcing the LFR technology, Dr. Clifford Reid, Complete Genomics’ chairman, president and CEO, said
We expect the introduction of this technological breakthrough to accelerate the move of whole genome sequencing into patient care, which in turn will begin to change the face of medicine.
Dr. Rade Drmanac, the company’s chief scientific officer and inventor of the LFR technology, said that
The Nature paper by Peters et al. describes how our LFR technology uses ‘barcoded’ DNA to generate whole genome sequencing with approximately one error in 10 million base pairs, or just 600 errors in an entire human genome. This represents a 10-fold increase in accuracy for Complete and is unmatched by any high-sensitivity method currently available.
In the not-too-distant future, failure to use phasing when providing genomic diagnoses in patient care will be seen as unacceptably inaccurate. I also suspect that LFR will reveal surprising things we didn’t know were missing because we didn’t have a tool to see them.