1. First, the MG workflow tries to find out if and where the pathogen exists among the 94 plates of input samples. To do so, it needs only 94 sets of reagents and reactions and equipment time. It will collect a fraction of the sample fluid in each well—called aliquot—from plate a3—the first plate—, add up aliquots from all 94 wells in the same input sample plate to form a pooled sample, concentrate it, purify it, and send for reaction (Fig. 1). Likewise for plate a4. Next, plate a5. And so on, until plate h12.
2. In practice, high throughput (HT) machines will run reactions on all 94 pools of samples concurrently in a plate of 96 wells (“Set 1 pooled sample plate”), with wells A3 through H12 holding the pooled samples from input sampling plates a3 through h12, respectively, while wells A1 and A2 holding positive/negative controls (Fig. 2, left half).
3. If all come up empty (aka negative), then that’s it—there’s no pathogen, all input samples are in the clear, and we move on.
4. But if there is one (or more) input sample that has the pathogen, one (or more) of the plates will turn up positive from the reactions. Then the method continues as follows.
5. Get a new 96-well plate. Leave the first two wells empty for positive/negative controls. Transfer aliquots from only the tested-positive input sample plate(s), with wells A3 to H12 receiving aliquots from the respective same-label wells of those plates.
6. In fast turnaround practice, setting up this new plate only after results from reactions on the Set 1 pooled sample plate are known adds unacceptable delay. In this situation, this new plate is set up and sent for reactions concurrently with the Set 1 pooled sample plate. In order to do so, instead of selecting only tested-positive input sample plates to pool aliquots from their wells, add up aliquots from all (94) same wells across all (94) input sample plates to place into each well of the new plate (“Set 2 pooled sample plate”) (Fig. 2, right half).
7. Take this Set 2 pooled sample plate through concentration, purification and reactions (Fig. 1). If none turns out positive, something is wrong with either this set of reactions or the previous one. Otherwise, at least one out of 94 will turn out positive.
8. Say exactly one plate turns positive, say plate e7, and exactly one reaction in the second set of reactions turns positive, say well D11. Then the pathogen has been found in input sample plate e7, its well D11. In this scenario, we have located the pathogen using 96*2 = 192 sets of reagents and reactions.
9. Alternatively, say two input sample plates turn positive, say plate e7 and h2, and two reactions in the second set of reactions turn positive, say well D11 and G8. Then the pathogen has been found those wells in those input sample plates, in total 4 wells: (1) input sample plate e7 in its wells D11 and G8, and (2) input sample plate h2 in its wells D11 and G8. Now take four aliquots from those wells from those two input sample plates and send for reactions, the third time now. It may turn out that all four test positive, or three of them, or two of them, but not just one. In this scenario, we have located the pathogen using 96*2 + 4 = 196 sets of reagents and reactions.
Concentrating and Purifying Pooled Samples: See protocols in uploaded pdf.
We are given the following:
At the above Steps 1 to 2 and 6, two pooled sample plates are produced—Set 1 and Set 2 pooled sample plates. Pooling 94 aliquots—250uL each—results in fairly large volume, almost 25mL—far more than can be held in a well—, and 94 of them.
Sample purification, here nucleic acid (RNA/DNA) extraction, using commercial kit, can only take in a small volume, 400uL for one commercial extraction kit (“sample volume”). At the end of extraction, 20uL is the smallest volume (“elution volume”) that is produced.
And finally, out of the elution volume, only 5uL will go into each reaction.
Two reactions will be performed per pooled sample. Test is positive if either turns positive.
How to concentrate 2 times 94 pools of nearly 25mL/pool down to 400uL/pool (about 60X) in a high throughput manner is the problem being addressed in this paper in order to make many-to-1 test methodology practical. The workflow must avoid human errors or cross contamination between samples or sample pools. This is the subject matter of the next section.
After concentrating the Set 1 and Set 2 sample pools, each set is laid out in 94 portions of 400uL/portion across 94 wells in a final pooled sample plate of 96 wells (wells A1 and A2 to be taken up by positive and negative controls). The plates are now ready to go for purification.
In sample purification, the two sets of 94 concentrates each go through conventional nucleic acid extraction to purify the pooled samples, using commercial high-throughput extractor machines and commercial high-throughput virus nucleic acid isolation kits.
Now, eluates (20uL each) from the purication step in two plates of 96 wells are ready for reactions. From each well, 5uL will be added to reagents for one reaction and another 5uL for a second reaction.