Skip to main content
Premium Trial:

Request an Annual Quote

Stanford Team Testing Whether 454 GS 20 Can Detect HIV Drug Resistance Better Than Sanger

Researchers from Stanford University have been testing whether 454’s sequencing platform can detect minor resistance variants of viruses involved in chronic diseases, including HIV and hepatitis B and C, that standard sequencing might miss.
Late last month they published a study in Genome Research exploring 454 sequencing for HIV resistance testing and suggested that massively parallel sequencing might eventually become the method of choice for characterizing drug resistance in chronic viral diseases and helping guide treatment.
“The 454 approach, or the more generic concept of ultradeep sequencing, is eminently logical for chronic viral diseases [in which] a person contains a range of different variants,” said Bob Shafer, an associate professor of medicine at Stanford. “So I think it’s only a matter of time [until] … this particular application, [or] something like 454, will be used.”
His team picked 454’s platform because it has been validated by many other researchers, and because its read lengths — 250 base pairs for the new FLX — are relatively long.
To be sure, “from the point of view of HIV drug-resistance [testing], the read length may not matter,” Shafer said, though it would enable researchers to study mutations that are linked in a virus. “However, researchers like myself generally are interested in longer read lengths because the data are more interesting” and could enable them to study the evolution of HIV populations in individuals.
At the moment, clinicians use Sanger sequencing to identify the 15 percent or so of HIV patients who will be resistant to existing drugs prior to treatment. But “when you sequence by standard sequencing methods, which are directly from a PCR product, you are likely to miss minor variants,” noted Shafer, who directs the genotypic antiretroviral resistance program at Stanford University Hospital.
But the researchers still need to overcome some technical hurdles to improve the quality and cost-effectiveness of sequencing clinical samples on 454’s platform. Also, clinical studies need to show that the information the assay yields is useful, according to Shafer. He guesses that a clinical assay is still five years away, “unless more groups joined and got involved in this, or we got more support” from either government agencies or diagnostic companies like Roche, which owns 454.
Others agree. What is needed to develop this into a clinically useful assay is “validation [that] it is a better predictor of response than standard assays” and a “reduction in cost,” John Mellors, a professor of medicine and chief of the division of infectious diseases at the University of Pittsburgh School of Medicine, wrote in an e-mail message.
In their Genome Research paper, Shafer, Mostafa Ronaghi from the Stanford Genome Technology Center, and their colleagues compared the ability of the GS 20 instrument and Sanger sequencing to characterize minor sequence variants in HIV-1 protease and reverse transcriptase genes from eight clinical samples.
Using a statistical method to distinguish real variants from sequencing errors, they found that 454 sequencing picked up about seven times more sequence variants on average than Sanger sequencing. Sanger sequencing of about 140 viral clones taken from one of the samples confirmed most of the minor variants detected by 454.
“That’s the novel aspect of our paper, demonstrating that despite the fact that the error rate [of the sequencing technology] is not insignificant, you can still, through appropriate statistical procedures, detect minor variants with very high accuracy,” Shafer said.

454’s single-read error rate “is still something I don’t want to sweep under the carpet; it’s still a problem.”

Nevertheless, the technology’s single-read error rate, which according to the article, is 0.1 percent on the GS20 for substitution errors, “is still something I don’t want to sweep under the carpet; it’s still a problem,” Shafer said, compared with the level of confidence obtained from standard sequencing. “And when you are dealing with clinical tests, that’s really important.”
False-positive sequences are indeed one of the major drawbacks of high-throughput sequencing for this application, Mellors agreed.
Increasing the single-read accuracy would also help to reduce the coverage needed to obtain a certain level of confidence, according to Shafer, and thus, to reduce the cost of the assay.
Cost could also be decreased by pooling and barcoding samples. Shafer is confident that his team will be able to cut the cost of detecting mutant variants at the 1-percent level to $500 per sample within the next six months or so, and possibly even push it into the $100 range “through a variety of both improvements of the technology as well as how we and others utilize the technology.”
By improving the extraction of virus RNA from a patient sample, the sequencing test could be made more sensitive as well.
According to Shafer, paying $500 or less per sample would enable him to perform larger retrospective studies, which would address, for example, what percentage of untreated patients harbor minor drug-resistance variants of the virus and how many of those fail treatment. “We are beginning to create collaborations to study that,” he said.
Since submitting their study for publication, the Stanford researchers have upgraded to 454’s FLX instrument and have sequenced additional HIV samples. They have also embarked on similar studies of HBV and HCV, including a collaboration with 454 on HBV.
They are not the only ones who are exploring massively parallel sequencing technologies for HIV drug resistance testing. Scientists from Yale University recently used 454’s FLX instrument to detect low-level drug-resistance variants in treatment-naïve HIV patients. Also, Intelligent Bio-Systems said it wants to apply its sequencing platform that relies on chemistry licensed from Jingyue Ju’s lab at Columbia University to test for HIV drug resistance (see In Sequence 6/19/2007)

The Scan

NFTs for Genome Sharing

Nature News writes that non-fungible tokens could be a way for people to profit from sharing genomic data.

Wastewater Warning System

Time magazine writes that cities and college campuses are monitoring sewage for SARS-CoV-2, an approach officials hope lasts beyond COVID-19.

Networks to Boost Surveillance

Scientific American writes that new organizations and networks aim to improve the ability of developing countries to conduct SARS-CoV-2 genomic surveillance.

Genome Biology Papers on Gastric Cancer Epimutations, BUTTERFLY, GUNC Tool

In Genome Biology this week: recurrent epigenetic mutations in gastric cancer, correction tool for unique molecular identifier-based assays, and more.