Scientific validation
The evidence
What the method measures, how well it measures it, and what it has found. Every figure on this page is redrawn from the published data and names its source. The papers themselves are the record, and both are open access.
01 · The method
Barcoding turns an immunoassay into a simple sequencing task
The barcoded MISPA™ library comes first: every antigen in it carries its own DNA barcode. One microlitre of sample is incubated with the whole library, antibodies bind their targets in solution where proteins keep the shape antibodies actually recognise, and the bound pairs are pulled down together. Counting the barcodes by sequencing is what lets one run cover thousands of samples.
02 · The library
One hundred and eighty-five antigens, measured against every sample
The panel used in the 2026 study spans the pathogens most people meet in a lifetime, plus the self-proteins that autoimmunity turns against. Every sample is measured against all of them. Adding targets increases the value of the library, but not the complexity of the assay.
What is in it
Coronaviruses including SARS-CoV-2 and the four endemic human coronaviruses; influenza and other respiratory viruses; the herpes family; enteroviruses and noroviruses; HIV, HPV and HTLV; bacterial species including streptococci, staphylococci, E. coli and Shigella; and autoantigens including p53, TRIM21 and interleukins.
Why full-length and folded
Antibodies recognise shape, not sequence alone. Peptide panels present fragments, which finds the antibodies that happen to bind a linear stretch and misses the rest. MISPA™ presents whole proteins in solution, so the targets look the way they look in the body.
03 · Validation
It agrees with the assays people already trust
Breadth is only worth having if each measurement is sound. Against qPCR-confirmed COVID-19 cases and pre-2019 samples, MISPA™ agreed closely with clinical diagnosis, and it agreed with FDA-cleared assays in the Frederick National Laboratory cohort.
Reproducibility
Across eight separate assay rounds, 81 control sera gave an average coefficient of variation of 18.1 percent and 192 blind samples 23.5 percent. In the method paper the average CV across all antigens was 23.7 percent, with the higher values where signals sit near zero, as expected.
Sensitivity
Detected antibodies down to 40 ng/ml, with a measured range up to 100,000 ng/ml.
04 · Scale
Eight thousand six hundred and eighty-seven samples, seven cohorts
The 2026 study followed 4,089 participants for up to eighteen months, from February 2021 to September 2022, across four centres in the SeroNet network: Mount Sinai, Arizona State, Minnesota and the Feinstein Institutes. Samples were taken before and after initial vaccination, and before and after the booster.
05 · Findings
Three things the breadth of MISPA™ made visible
Most antibody levels barely move, which is what makes change meaningful
In the 743 participants who gave three or more samples across eighteen months, antibody levels against nearly every antigen held steady. The exception is the one that should move: SARS-CoV-2, during a vaccination campaign. That stability is the reason a change in an individual can be read as an event rather than as noise.
Vaccine response is far more variable than the schedule assumes
Anti-RBD levels were measured before and after the initial two doses in 1,569 people, and before and after the booster in 2,412. After the initial course, 72 percent of transplant recipients and 48 percent of myeloma patients had a weak or negative response. Boosting closed much of that gap, but not all of it: 38 percent of transplant recipients were still weak or negative afterwards. Six percent of people who reported themselves healthy responded weakly to the initial course.
Antibodies you already carry predict how you will respond to a new vaccine
Models built on antibody responses to common microbes, with SARS-CoV-2 and HIV excluded, separated strong from weak booster responders using only pre-booster samples. The signal is strongest in inflammatory bowel disease and holds above chance in every cohort tested.
06 · Papers
Peer-reviewed, and open
Two papers describe the method and what it found. Both are open access, and they are the record. Everything above is drawn from them.
Pre-vaccine sentinel antibodies predict blunted vaccine responses
Antibodies a person already carries against common microbes predict how strongly they will respond to a new vaccine. In 4,089 people, 1,644 of them immunosuppressed, MISPA™ measured 185 antigens before and after COVID-19 vaccination and booster. Models built on those profiles separated strong from blunted responders, and blunted responses appeared in healthy people as well as immunosuppressed ones.
Quantitative assessment of multiple pathogen exposure and immune dynamics at scale
The method paper. Responses to 39 bacterial species and strains and 99 viruses were measured in 2,400 people. The assay is quantitative over a wide range, uses full-length folded proteins, and performed favourably against commercial clinical assays. Antibody levels were stable across the six-month window of that study, so a change in level marks an intervening event. The 2026 study extended the same finding to eighteen months.
Intellectual property
Genome-scale detection of antibody responses using mRNA display and next-generation sequencing
US 9,938,523 · 10,717,977 · 11,739,314 · 12,030,909Nucleic acid-tagged compositions for multiplexed protein–protein interaction profiling
US 12,180,467Dual barcode indexes for multiplex sequencing of samples screened with MISPA™
PCT/US21/042784MISPA™ is a research method described in peer-reviewed literature and used in research settings. It is not a cleared or approved clinical diagnostic test, and nothing here should be read as a diagnostic claim.
Collaborate
Start with a question.
End with insight.
Two things start a collaboration. Everything else follows from them.