The Signal Report // LAB SCIENCE
Lab Science.
The methods behind the headline result, and where the press release outran the paper. All reports →
16 in this beat · page 1 of 2
The Contamination You Cannot See
Mycoplasma leaves the medium clear, has no cell wall for penicillin to attack, and passes standard filters. Three safety nets, none of them catching it. Then two researchers found it by reading RNA-seq data that hundreds of labs had already deposited in a public archive for entirely unrelated reasons.
Read the report → Report 098 · Lab Science · 2026-08-20 · 8 minThe Edge Wells Are Not the Same Experiment
Every cell biologist is told to skip the outer ring of a 96-well plate. One group measured what is actually happening: 35 percent lower signal in the outer wells, reaching three rows in, with the plate brand mattering more than the technique. A second study found the effect running the other way.
Read the report → Report 092 · Lab Science · 2026-08-17 · 10 minThe $38 Trillion Number That Was Retracted
It was the second most media-covered climate paper of its year, central banks cited it, and its own authors withdrew it in December 2025. The central estimate barely moved; the confidence around it moved a great deal. Underneath sits an error with a name, one that biology students are taught and almost nobody checks for: twenty regions of one country are not twenty independent facts about the world.
Read the report → Report 086 · Lab Science · 2026-08-14 · 8 minFive Levels of Identified by Mass Spec
A mass spectrometer measures masses, not identities. Analytical chemistry grades the gap on a five-level scale, and only level 1 requires that somebody ran the authentic reference standard. Two competing scales, one shared trap, and why even level 1 can leave you with two molecules.
Read the report → Report 080 · Lab Science · 2026-08-11 · 9 minWhat NIST Traceable Actually Buys You
The bottle your instrument is calibrated against carries its own uncertainty, on the order of a couple of percent for some elements, with a coverage factor printed per analyte because it is not always 2. Its certification expires on a date, is nullified by improper storage, and NIST's own policy says establishing the traceability of your result is your job, not theirs.
Read the report → Report 074 · Lab Science · 2026-08-08 · 10 minReproducible and Replicable Are Not the Same
Nature published a package of meta-research on 1 April 2026. Two papers landed on a number near half and the coverage merged them. They measured different failures, and the more basic one could only be tested on a quarter of the sample.
Read the report → Report 068 · Lab Science · 2026-08-05 · 10 minWhy the Legal Limit Is 4 Parts Per Trillion
The health goal for PFOA in drinking water is zero. The enforceable limit is 4.0 parts per trillion. EPA's own rulemaking record explains the gap, and the answer is not cost or treatment technology. It is that 4.0 is the lowest concentration three quarters of American laboratories can reliably measure.
Read the report → Report 062 · Lab Science · 2026-08-02 · 9 minWhat "Not Detected" Actually Means
The EPA's binding definition of a detection limit compares your sample not to zero but to the laboratory's own blanks, established over three calendar dates on one instrument. So the limit is a property of the lab. Meanwhile the health goal for PFOA in drinking water is zero and the enforceable limit is 4.0 parts per trillion, and both numbers are deliberate.
Read the report → Report 056 · Lab Science · 2026-07-30 · 8 minWhat Those Error Bars Actually Mean
Standard deviation, standard error and a 95% confidence interval get drawn as the same little vertical line, and they answer completely different questions. One of them shrinks automatically as you add data. Worse is the failure underneath: counting 300 cells from one dish as n = 300 manufactures confidence that was never in the experiment.
Read the report → Report 050 · Lab Science · 2026-07-27 · 7 minAre Your Cells What the Label Says?
A testing lab published what came through its door: of the human cell lines sent in for identity checks in 2024 and 2025, a few percent were not the line on the vial. Then the authors wrote one sentence about their own sample that turns the number into a floor. A measurement scientist on the one measurement nobody thinks is a measurement.
Read the report →