NEW YORK (GenomeWeb) – In an effort to understand variability among microbiome studies, researchers at Human Longevity and the University of California, San Diego recently compared four library preparation products for whole-genome sequencing.
Specifically, the group quantified the biases associated with four library prep methods for the Illumina HiSeq 2500 platform using both a synthetic mock community of 20 microbes and an extraction control of microbial cells spiked into stool samples. The researchers compared the performances of the Illumina TruSeq DNA PCR-free and Kapa Hyper Prep PCR-free methods to the Illumina Nextera XT and Kapa Hyper Prep PCR kits.
Measures of relative abundance of microbes in the mock community varied among the methods, but the two that did not rely on PCR yielded lower error rates and higher-quality reads.
In results published online last week in the Proceedings of the National Academy of Sciences, the researchers therefore concluded that PCR-free methodologies should be used for library prep in order to reduce PCR bias.
There are multiple points where bias could be introduced in next-generation sequencing, such as library prep and size selection, the clustering reaction, and sequencing, but, "by switching to a PCR-free system, we remove a huge potential for that bias," Marcus Jones, a researcher at HLI and first author on the study, told GenomeWeb in an interview.
Jones said the evaluation was done to vet and validate the methods for HLI's own pipeline. The company offers various products and services for genomics and cell-based diagnostic and therapy development, and was recently CLIA certified, as previously reported.
The mock microbial community was characterized using "gold standard" qPCR methods. "We used three different genes that are all single copy in each of the genomes, and we looked across to make sure we were getting reproducibility in the abundance measurement," Jones explained.
None of the four methods precisely duplicated the qPCR proportions, but they were within two-fold for most microbes. "The level of quantitative variability of [relative genome abundance] compared with qPCR is extremely interesting and alarming," the authors noted in the study, since, "Numerous publications report a change in abundance of 0.1 percent as significant."
There could be a number of explanations for the superiority of the PCR-free methods. There are certain error rates in PCR, which may be polymerase specific, and the different PCR-based systems require different numbers of PCR cycles, Jones noted.
There are also inherent biases in PCR, and increased PCR duplication rates occur when you lose complexity of a sample.
Finally, the Nextera method uses tagmentation notation, which is enzymatic fragmentation of the DNA that may also introduce bias, while the Kapa and TruSeq PCR-free methods use mechanical sonication of DNA, Jones explained.
Standardizing library prep may become more topical after recent calls for improvements in microbiome research, as reported by GenomeWeb.
But Jones said it will likely not be appropriate for researchers to recommend a single system, although guidelines could be put in place.
"I think what we would do is make strong recommendations on how samples should be collected, how they should be processed, and the best practices, not necessarily [endorsing] a particular company, but say, if there is enough material, use a PCR-free system; if not, use a low-cycle PCR-based system," Jones said.
He could not disclose which method HLI will use going forward, but he noted that the company tailors its approaches to the sample types. The group has no financial interests with Kapa or its parent company, Roche, which acquired the firm a few months ago, but "the Kapa Hyper Prep looks very appealing [because] it gives you the option of doing PCR-free or low-cycle PCR."