My Blood Test Results...

I’m not dying or anything¹, I just thought it would be fun to share…

Every few months I go into Chiba University Hospital and receive medication for nothing particularly serious. As part of this they extract some blood and urine from me and process it into a report… here’s one of these reports:

A doctor then looks at this report. Specifically they look at the little “L”s and “H”s. They then ignore them or they say something vague that I interpret as “Eat Less Pizza”.

What are these tests? What instrumentation is used? And can we theoretically, in the far future, replace any of them with sequencers?

The tests are broken down roughly as follows:

• Metabolic Tests

• Blood Tests

• Antinuclear Antibody Tests

Metabolic Tests

These are the tests that tell me if I’ve been eating too much Pizza. Cholesterol levels, Sodium levels… those sorts of things (obviously a number of other non-Pizza related analytes are included).

I can’t see any plausible way these could be replaced by nucleic acid sequencing.

Blood Tests

These are various tests which as I understand it would normally be described as complete blood counts.

Cell Counts

The fundamental tests are red and white blood cell counts. These counts can be done with a “simple” microscope slide:

Slide prep from here.

More commonly they now use various kinds of hematology analyzers, using coulter or optical cell counting. Could you do red and white blood cell counts with sequencing?

I don’t know of anyone who has seriously considered this. I suspect you could get a reasonable white blood cell count based on DNA concentration:

From here.

It seems possible you could use sequencing here, possibly with a spike-in for calibration.

Red blood cell counts seem less obviously addressable using sequencing. Red blood cells lose their nucleuses during development… but it seems they retain a bunch of RNA and the literature around red blood cell transcriptomics seems quite interesting. 2 Similar platelet sequencing seems like it might be theoretically viable.

Doing it with Nucleic Acid Sequencing…

I suspect you’d be hard pressed to replicate red and white blood cell counts with whole blood sequencing alone. You’d perhaps have a better chance with single cell sequencing. I suspect you’d need to do RNAseq.

We’re talking days and hundreds of dollars per sample. You can compare this to a Horiba blood analyzer which process up to 15+ tests an hour:

ANA Tests

Looks like I also get “Antinuclear Antibodies” tests. Here they’re looking for autoantibodies, evidence that my immune system is attacking healthy cells.

ANA Tests can be performed in a variety of ways, including using ELISA. The way it’s reported here makes me wonder if they’re not just looking at a microscope slide…

From here.

Not clear to me there’s any viable way to address these tests using nucleic acid sequencing…

30 Years Into The Future

What if we look 30 years into the future and assume cheap, instant sequencing is widely available. Do we then have a route to replacing some or all of these tests with a DNA sequencer?

This seems very unlikely.

The problem is that even the ideal sequencer only a small subset of the tests are even theoretically addressable with nucleic acid sequencing. Specifically none of the following:

• The Biochemical/metabolic tests

• The blood hemoglobin tests.

• The CRP (inflammation related protein) measurements.

• The Antinuclear antibody tests

So… what? Why did would we want to use sequencing anyway?

If we don’t use sequencing as a first line test, it gets pushed further down the diagnostic flowchart. We identify pathogenic causes later or not at all. On a societal level that means catching pandemics when they’ve spread beyond any hope of containment.

On an individual level that obviously means waiting longer to figure out what’s wrong…²

However… I still haven’t entirely given up on the idea of a StarTrek-style medical tricorder.

Perhaps instead of nucleic acids we need to look to protein sequencing instead. It seems to me that protein sequencing could at least theoretically address all but the biochemical/metabolic tests.

So I’ll be pinning my hopes on hobbling my way into Chiba University Hospital in 2054 to have my blood and urine run through a single cell combined DNA/RNA/Protein sequencer, which costs ~$10 per run and delivers billions of data points.

It’s a pretty wild dream… but we can dream.

Subscribe for more dreams (and more serious posts)

1. Obviously we’re all dying on some timeline… but I have no immediate plans.

2. Or never finding out, and certainly missing signals which might be aggregated on a societal level to identify novel pathogens.