Everything

How the list was made

We searched Europe PMC, Crossref, NASA ADS and publisher sites for papers that report resistive pulse measurements, starting with Coulter's patent and working forward to 2026. For each paper we recorded what was measured, the pore or aperture, how the pore was made, the run conditions and, when the paper gives them, the spike depth, the dwell time and the event rate. Claude, an AI model made by Anthropic, ran the searches and pulled the details from each source, and every entry keeps a link to the source it came from.

We marked each entry with how we checked it. "Full text" means the details came from the paper itself, "abstract" means from its abstract, and "described in a later paper" means a review or a later study described the work (common for papers from the 1950s and 1960s, which rarely have abstracts online). We checked every DOI against Crossref (or, for the few DOIs that Crossref does not hold, against the DOI registry) and fixed or removed the ones that pointed at the wrong paper.

We had other AI agents, working apart from the ones that built the list, check random entries against their papers in two rounds. In the first round of 55 entries, every entry pointed at the right paper, and 54 of the 55 papers did measure that analyte by resistive pulse sensing. The 55th paper used a laser instrument that carries the Coulter brand name, and we removed it along with 14 more entries from the same two papers. Most of the other errors were pore details, such as a standard pore size or "planar lipid bilayer", that an entry stated but its paper did not. A script then compared every entry with the text of its source, where Europe PMC had that text, and removed the pore details and numbers that the text does not contain. In the second round of 40 entries, every paper and every measurement checked out, and 17 entries still had a slip in one field, most often a number that differs a little from the paper. We fixed every slip the checks found. Check the paper before you rely on a number.

What counts as one analyte

Every different molecule, cell or particle counts once. Two DNA strands with different sequences are two analytes, and so are two proteins that differ by one amino acid, the two mirror forms of one drug, or red cells from two species. Beads of one material that differ only in size count as one analyte, with their sizes listed. An analyte measured in many papers appears once, with its measurements listed under it: the first report we found, plus later reports that used another kind of pore or said more about the spikes.

Where the sizes, masses and charges come from

Sizes, masses and charges come from the papers when the papers report them. For DNA and RNA with a known sequence or length, we computed the mass and the charge of the phosphate backbone. For proteins with a UniProt entry, we computed the mass, the isoelectric point and the net charge at pH 7 from the sequence. For small molecules and ions, the mass, formula and charge come from PubChem. For cells, microbes, viruses and particles, we give the charge as a zeta potential or an isoelectric point when a source reports one. We give the source next to each value.

What the list leaves out

Nanopore sequencers have read DNA and RNA from many thousands of organisms, and the list gives the first reports and the landmark samples. Hematology analyzers count blood cells in a huge share of clinical papers, and the list gives each kind of cell once or a few times. Some measurements surely remain unfound, most of all in old industrial and food journals that have no abstracts online.