DNA nanoball sequencing

Cells are lysed and DNA is extracted from the cell lysate. The high-molecular-weight DNA, often several megabase pairs long, is fragmented by physical or enzymatic methods to break the DNA double-strands at random intervals. Bioinformatic mapping of the sequencing reads is most efficient when the sample DNA contains a narrow length range.[7] For small RNA sequencing, selection of the ideal fragment lengths for sequencing is performed by gel electrophoresis;[8] for sequencing of larger fragments, DNA fragments are separated by bead-based size selection.[9]

Adapter DNA sequences must be attached to the unknown DNA fragment so that DNA segments with known sequences flank the unknown DNA. In the first round of adapter ligation, right (Ad153_right) and left (Ad153_left) adapters are attached to the right and left flanks of the fragmented DNA, and the DNA is amplified by PCR. A splint oligo then hybridizes to the ends of the fragments which are ligated to form a circle. An exonuclease is added to remove all remaining linear single-stranded and double-stranded DNA products. The result is a completed circular DNA template.[2]

Once a single-stranded circular DNA template is created, containing sample DNA that is ligated to two unique adapter sequences has been generated, the full sequence is amplified into a long string of DNA. This is accomplished by rolling circle replication with the Phi 29 DNA polymerase which binds and replicates the DNA template. The newly synthesized strand is released from the circular template, resulting in a long single-stranded DNA comprising several head-to-tail copies of the circular template.[10] The resulting nanoparticle self-assembles into a tight ball of DNA approximately 300 nanometers (nm) across. Nanoballs remain separated from each other because they are negatively charged naturally repel each other, reducing any tangling between different single stranded DNA lengths.[2]

DNA nanoball creation and adsorption to the patterned array flowcell

To obtain DNA sequence, the DNA nanoballs are attached to a patterned array flow cell. The flow cell is a silicon wafer coated with silicon dioxide, titanium, hexamethyldisilazane (HMDS), and a photoresist material. The DNA nanoballs are added to the flow cell and selectively bind to the positively-charged aminosilane in a highly ordered pattern, allowing a very high density of DNA nanoballs to be sequenced.[2][11]

After each DNA nucleotide incorporation step, the flow cell is imaged to determine which nucleotide base bound to the DNA nanoball. The fluorophore is excited with a laser that excites specific wavelengths of light. The emission of fluorescence from each DNA nanoball is captured on a high resolution CCD camera. The image is then processed to remove background noise and assess the intensity of each point. The color of each DNA nanoball corresponds to a base at the interrogative position and a computer records the base position information.[2]

The data generated from the DNA nanoballs is formatted as standard FASTQ formatted files with contiguous bases (no gaps). These files can be used in any data analysis pipeline that is configured to read single-end or paired-end FASTQ files.

For example:

Read 1, from a 100bp paired end run from[12]

 @CL100011513L1C001R013_126365/1
 CTAGGCAACTATAGGTCTCAGTTAAGTCAAATAAAATTCACATCAAATTTTTACTCCCACCATCCCAACACTTTCCTGCCTGGCATATGCCGTGTCTGCC
 +
 FFFFFFFFFFFGFGFFFFFF;FFFFFFFGFGFGFFFFFF;FFFFGFGFGFFEFFFFFEDGFDFF@FCFGFGCFFFFFEFFEGDFDFFFFFGDAFFEFGFF

Corresponding Read 2:

 @CL100011513L1C001R013_126365/2
 TGTCTACCATATTCTACATTCCACACTCGGTGAGGGAAGGTAGGCACATAAAGCAATGGCAGTACGGTGTAATACATGCTAATGTAGAGTAAGCACTCAG
 +
 3E9E<ADEBB:D>E?FD<<@EFE>>ECEF5CE:B6E:CEE?6B>B+@??31/FD:0?@:E9<3FE2/A:/8>9CB&=E<7:-+>;29:7+/5D9)?5F/:

Default parameters for the popular aligners are sufficient.

In the FASTQ file created by BGI/MGI sequencers using DNA nanoballs on a patterned array flowcell, the read names look like this:

Anatomy of a BGI sequencer read name

Anatomy of an MGI sequencer read name

BGISEQ-500: CL100025298L1C002R050_244547

MGISEQ-2000: V100006430L1C001R018613883

Read names can be parsed to extract three variables describing the physical location of the read on the patterned array: (1) tile/region, (2) x coordinate, and (3) y coordinate. Note that, due to the order of these variables, these read names cannot be natively parsed by Picard MarkDuplicates in order to identify optical duplicates. However, as there are none on this platform, this poses no problem to Picard-based data analysis.

Because DNA nanoballs remain confined their spots on the patterned array there are no optical duplicates to contend with during bioinformatics analysis of sequencing reads. It is suggested to run Picard MarkDuplicates as follows:

java -jar picard.jar MarkDuplicates I=input.bam O=marked_duplicates.bam M=marked_dup_metrics.txt READ_NAME_REGEX=null

A test with Picard-friendly, reformatted read names demonstrates the absence of this class of duplicate read:

Test of Picard MarkDuplicates varying the OPTICAL_DUPLICATE_PIXEL_DISTANCE parameter

The single read marked as an optical duplicate is most assuredly artefactual. In any case, the effect on the estimated library size is negligible.