Episode Summary

Charles Chiu, MD, PhD, UCSFMost diagnostic tests look for a single microorganism, or at most a limited panel of microorganisms. Charles Chiu discusses his research on metagenomic sequencing as a diagnostic tool that can identify all potential pathogens in a given patient sample.

Subscribe (free) on Apple Podcasts, Google Podcasts, Android, RSS, or by email. Also available on the ASM Podcast Network app.

Julie’s Biggest Takeaways

Many patients have infectious diseases that go undiagnosed, despite the best efforts of clinical microbiologists. This is in part due to:

  • Indistinguishable clinical manifestations for diseases such as meningitis, respiratory infections, and sepsis, when caused by different microorganisms.
  • Insuffient sample to run all specific tests.
  • Insufficient resources to send samples to clinics that can make specific diagnoses.
  • Limitations to conventional tests such as culture are limited for microorganisms that don’t culture well.

One of the key factors that determines just how much microorganism genetic material needs to be present relative to host genetic material is sequencing depth for a given test. Greater depth of sequencing improves sensitivity and ability to detect low-concentration samples.

Microorganism type also factors into applications of metagenomic sequencing as a diagnostic tool. Staphylococcus aureus and Staphylococcus epidermidis genomes are more than 99% identical, and therefore require many sequencing reads to differentiate the two species. But because viral genomic sequences are highly specific, 3 unique reads of a viral genome (from different regions of the genome) are sufficient to confidently diagnose a viral infection.

Two key improvements have improved metagenomic sequencing to the point that it can be incorporated into clincal microbiology labs:

  • Improved sequencing throughput. Original Roche 454 sequencers allowed around 100,000 reads per run, but NovaSeq machines can generate up to 1 billion sequences in a run. This improves sensitivity of detection and facilitates sample multiplexing.
  • Development of bioinformatics pipelines. Using BLAST-related software to analyze the high number of sequencing reads would take days; new systems allow the analysis turnaround time for millions of reads to be completed in 1-2 hours. This is essential for diagnosing critically ill patients.

Diagnostic test types currently available:

  • Culture: Directly identifying the microorganism.
  • Serology: Indirectely identify patient response to the presence of a microorganisms.
  • Antigen-based methods: Directly identify pieces of the microorganism.
  • Molecular-based methods: Directly detect nucleic acids associated with the microorganism. However, PCR-based methods require that you know the sequence for which you’re looking.
  • Metagenomic sequening: allows interrogation of all potential microorganisms that may be present, and can be enriched with the addition of probes for certain pathogens or microorganism types.

Is metagenomic sequencing likely to displace existing tests? No. It is a direct detection test that relies on nucleic acids being present in the sample and thus may not diagnose some infectious etiologies. For example, infections where the microorganism is present only transiently may not leave nucleic acids that can be detected by the time symptoms begin to show. Charles sees metagenomic sequencing as complementary to currently available tests that will add to the number of ways diseases can be diagnosed.

Featured Quotes

“One thing that has always been challenging in clinical practice have been the large number of patients who are admitted to hospitals where we’re unable to provide a diagnosis for their infection, and this is despite extensive conventional testing. It’s really been my dream that we can potentially bring in new technologies like metagenomic sequencing as a way we can increase the number of diagnoses in these critically ill patients.”

“The key difference in metagenomic sequencing for infections is that we’re looking at the nonhuman fraction of sequence reads, which tends to be very, very small. It can be less than 0.1% of your data.”

“I think in the next 5-10 years, you’re going to see vast improvements in sequencing technologies that enable you to do real-time analysis...this really opens up a huge number of possibilities.”

“Metagenomic sequencing is not available to a number of laboratories, for reasons which typically relate to the cost of implementation as well as the amount of infrastructure that’s needed for this technology to be implemented.”

“Metagenomic sequencing is effectively doing a microbiome experiment. We now know that infections don’t occur in isolation; the traditional thinking has been one bug, one disease. I think that’s largely changed and we know that infectious disease is a complex interplay between the human host, between colonizing microorganisms, and between pathogens. Being able to tease that out by doing both comprehensive microbiome profiling of, say, respiratory secretion of a patient with pneumonia: I think that’s going to yield a lot of diagnostic insight that may be leveraged into a clinical test someday.”

“By introducing these new technologies, I feel we’re really opening up the next generation of microbiologists, where we arguably no longer have to rely on traditional or arguably outdated methods of diagnosing disease.”

Links for This Episode

History of Micobiology Tidbit


When did doctors start to think about point-of-care tests? One can argue it was suggested over 100 years ago, when immunoassays were used but antibodies and their relationship to clonal selection and development of specialty antibody-secreting plasma cells hadn’t yet been uncovered.


Alphose Dochez and Oswald Avery were studying the pneumococcus in the early 1900s. Streptococcus pneumoniae can cause a number of serious diseases, and even today pneumococcal pneumonia has a mortality rate of 5-7%. 100 years ago, before antibiotics were widely used, mortality was closer to 30-40%. Dochez and Avery were studying a soluble substance secreted into the medium during bacterial growth - and they describe the purification of the ‘soluble bacterial substance’ from in vitro cultures. Despite not knowing that this substance was capsular polysaccharide, the scientists hypothesized that detecting this substance in animals would work to determine which animals were infected.

They tested this hypothesis first with the body fluids from infected rabbits. Filtering to remove bacteria, the serum and urine were tested for the presence of soluble bacterial substance using the same protocol. Finding this precipitable substance in animals, they then show the same substance can be purified from the blood of patients suffering from pneumonia. In testing 25 patients, they found the substance in either blood or urine, depending in part on the pneumococcal type, or serotype, a category based on interaction of the immune system with the capsule.

In their discussion, Dochez and Avery note that the amount of precipitable substance in the urine seems to be a measure of the severity of the infection, and that a high amount of substance in the urine correlates with a high mortality rate - giving the precipitin test in urine a “considerable prognostic value.” They further suggest that the test “may also be used in making a rapid diagnosis of the type of organisms with which an individual is infected,” which was important for administering the correct antisera, the preferred therapy at that time. A very prescient observation, especially given that they weren’t sure what the soluble bacterial substance was!

If the names of these scientists and the bacterium sound familiar, that may be because Avery’s experiments on capsule and transformation of smooth and rough colonies of Streptococcus pneumoniae is one of the major experiments demonstrating that DNA is the molecule of heritability. I’m constantly in awe of the acheivements of the early molecular biologists, who were truly renaissaince scientists working on multiple important projects simultaneously.

Let us know what you thought about this episode by tweeting at us @ASMicrobiology or leaving a comment on facebook.com/asmfan.

Send your stories about our guests and/or your comments to jwolf@asmusa.org.