Dr. Norman G. Anderson’s comments on Dr. Michael Montague’s post of August 30:
Global Screening for Human Viral Pathogens has been described in detail in a CDC publication
Briefly the procedure includes nationwide routine collection of both normal and diagnostic sera, shipment of these to centralized laboratories for analysis, pooling of excess diagnostic sera now routinely discarded (amounting to hundreds of liters per week), recovery of the viral loads from 100 liter batches using the K-II centrifuge (which I developed in collaboration with the Gas Cenrifuge program at Oak Ridge), further concentration and purification using microgradient ultracentrifuges recently developed, shotgun sequencing of the concentrates, and data reduction and sequence reconstruction.
Assuming 1 ml per original donor, and ten centrifuges in parallel, one million cases can be scanned per run, assuming sufficient collection. The total viral load is concentrated down to less than the average volume of one original sample, hence the final concentration can approach that in each original viremic donor. In general viremia peaks with fever, hence the importance of obtaining samples on admission, as can be done here. The methods apply equally well to urine, and modifications apply to tissues. A complete concentration run takes less than three days. The objective is a running inventory of all viruses in circulation in man.
The K-II centrifuge is used world-wide to purify commercial vaccines, especially those for influenza. Much of this work depends on studies done on marine viruses. Organization of this work is proceeding in China.
Dr. Norman G. Anderson: President, and Founder of the Viral Defense Foundation. Atomic Energy Commission Citation and Gold Medal for the invention and development of the K-II vaccine-purification ultracentrifuge, 1972. Over 250 publications and 44 issued patents.
Biosecurity Blog 
The centrifuge based strategy that Dr. Anderson proposes is an excellent example of the sort of technique that would be used in the third stage of Michael Montague’s previously described plan Toward Continuous Real-Time Bio-Surveillance .
Why the third stage of Michael Montague’s plan and not the first? The barriers for creating a national bio-surveillance system that reaches a representative selection of the US population are considerable and varied, and by-enlarge, not bio-technological in nature.
The first and most severe such barrier is public fear which in turn drives policy. The US is a nation that has so far resisted even the elementary step of a unified national ID card even to the point of States refusing to cooperate with the federal Real ID Act. Video surveillance systems that would be introduced in other parts of the developed world with little or no thought for privacy are strongly resisted in the US. This strong distrust by the public of anything that could be called “government surveillance” remains very alive even after 911. One of the many consequences of this fear is the creation of government policies to address it. There are a large number of laws and regulations, including but not limited to HIPAA, that impose complex and stringent regulation concerning the handling of patient data and samples. Pooling samples and getting a read-out of viruses resident in a clinical population might have been legal in 1970’s, when such technology was being developed at Oak Ridge, but is now something that would, by law, require significant administrative oversight. Such oversight would need to be in place before, not after, the technology-component was in place otherwise fear-mongers on the grass-roots and political levels would kill such a bio-surveillance effort before it could begin. This is one of the reasons that Michael Montague’s previously stated plan does not start with the introduction of new or old technology, but rather with an administrative framework using only data that is already being collected.
Another barrier to creating a useful bio-surveillance program is data management and analysis. When was a sample taken? When was it analyzed? When was the related data transmitted to a central database? Where is it from? Is that the location of the patient’s home? Of the hospital? Was the analysis performed at one of those locations, or at a third location? Each of these questions alone, and for a single specimen, are easy. However, managing and characterizing the answers to all of these questions at the same time, and for thousands to millions of specimens, and doing so without causing privacy concerns, see above, and doing it in anything like a real-time manner is NOT easy. This is the other reason why Michael Montague’s plan focuses upon building a central database using existing data first, and creating new forms of data to populate it only later. Without a firm handle on the meta-data associated with a pathogen detection, such a detection is all but useless.
This explains why would Dr. Anderson’s pooling and centrifuging technology can not be the the first stages of a bio-surveillance program, but why shouldn’t it be the last? In essence, Dr. Anderson’s proposal is a metagenomic study of the viral fraction of pools of clinical samples. As such, we can expect all of the traditional difficulties that have been observed with other metagenomic studies:
Detection by sequencing suffers from numerous problems. Even with immuno and hybridization based subtraction of known viruses, many rounds of sequencing will be required to achieve anything like a complete characterization of the pooled sample. Without such subtraction, the most populous species in the population will be sequenced many times consuming output bandwidth. Consumption of bandwidth prevents rare species in the pool from being sequenced at all. Next generation sequencing technology has not been able to add enough bandwidth to directly overcome this. Immuno-subtaction, and subtractive-hybridization could eventually overcome this, but only with multiple rounds of sequencing. Initial rounds of sequencing are needed to determine the most populous viruses in the pool. Only then would one know which viruses to subtract out from subsequent rounds of sequencing. Multiple rounds of sequencing followed by subtraction by antibodies and/or hybridization can rapidly become extremely expensive, and is always very slow because libraries must be remade. The result is that, after the centrifuge step, each pool would take many weeks or even months to analyze (and that’s without any effort to assemble the resulting reads). Thus, with current methods, data-sets from pooled samples analyzed by sequencing, would require vast resources to be available in anything like real-time. With the removal of the sequencing detection method and the introduction of a multiplex PCR coupled with microarrays, and/or an ELISA-on-a-chip detection method in it’s place, something like Dr. Anderson’s centrifuge purification of the viral fraction of pooled samples would be quite useful. This is exactly the sort of thing described in the third stage of Michael Montague’s plan. Of course, not all biological threats are viral, so centrifuge based purification could only be one part of a larger detection strategy even at the third stage of the plan.
However, from an epidemiological point of view, pooled samples provide an imperfect data-set regardless of the technological details of how they are analyzed. This is because the viral load of a given infection is highly variable from patient to patient and at different points during the infection. The consequence is that there is no definitive way to distinguish between 1% off the contributors to a pool being positive for a given pathogen, and 100% being positive or anything in between. Also pooling samples imposes a timing penalty since the oldest sample to be added to a pool is delayed waiting for the last sample. For periodic but infrequent snap-shots of the circulating pathogens this might not be a major concern, however, the detection of a natural epidemic or a biological attack is not something that can wait. Look at how rapidly the swine flu situation developed. This is why Michael Montague’s plan does not stop at just pooled samples (the third stage of the plan), but proceeds toward a larger goal of cheap, fast, individual, broad-spectrum detection with anonymous national data gathering and analysis.