Hacker News new | ask | show | jobs
by adrianmonk 1276 days ago
> What is PMMoV normalization?

> In addition to the pathogens it tests for, WastewaterSCAN measures an extremely common, harmless plant virus that is consumed when people eat. It is called pepper mild mottle virus (or PMMoV). By measuring the concentration of PMMoV genetic markers per dry weight gram of wastewater solids, WastewaterSCAN can account for how much viral material is recovered from each sample and changes to the “fecal strength” of that sample. For example, heavy rain that drains into a wastewater system can dilute the strength of a particular day’s sample.

OK, that's pretty clever and it sounds useful, so I guess I should turn on that toggle. But then I read further and see this:

> Our analysis suggests that both the concentrations of the SARS-CoV-2 genes measured as copies per gram of solid waste and those concentrations normalized by PMMoV are proportional to laboratory confirmed COVID-19 incidence rates in the sewersheds.

How can both things be proportional? Do they actually mean positively correlated?

If raw data is already proportional to lab tests, then it doesn't seem like there's any benefit from normalizing.

3 comments

I think a lot is being lost in the translation here. I mean, Verily's analysis is a laboratory *determined* COVID-19 incidence rate in the sewersheds right? How is it any different from the laboratory *confirmed* incidence rate they compare to? As far as proportional, I think you are right. When laboratory confirmed goes up, both Verily concentrations also go up. When laboratory confirmed goes down, both Verily concentrations also go down. I doubt there is a constant scale factor operating here unless the laboratory confirmed is using a similar method of normalization (i.e., they all start with copies per gram dry weight).
Wastewater analysis is a new and very difficult measurement problem. There's continual discussion of how to do the analysis, and no strong consensus in the field.

This is the best part of science: not knowing, but working to discover.

The strength will mostly be a constant scale factor, hence proportional to.