Jones — PFAS Analytical Chemistry & blueFish

Emeritus domain expert providing the product specification and validation rubric for blueFish through decades of analytical chemistry expertise.

Status: Active — emeritus consulting confirmed, technical resources received Domain: PFAS analytical chemistry, mass spectrometry, environmental monitoring Product: blueFish (sovereign PFAS analytical chemistry ETL) Springs Fed: wetSpring Track 2


The Collaboration

This is not a typical gen5 collaboration (PI + grant). It is an emeritus domain expert providing the product specification and validation rubric for blueFish through accumulated institutional knowledge — 30+ years of analytical chemistry domain expertise.


What Domain Expertise Revealed

EPA Method 1633A Gap

Most labs (including commercial) do not actually perform all QC checks — particularly ion abundance ratios. No published data exists to evaluate QC compliance beyond single-letter qualifiers in lab reports. blueFish implements full 1633A QC including ion abundance ratios — the gap commercial labs are not filling.

Open-Source Tool Assessment

ToolAssessment
MZmineUnacceptable failure rates
MS-DIALUnacceptable failure rates
XCMSUnacceptable failure rates
Progenesis QIGood graphical evaluation, now obsolete

No open-source tool gives acceptable failure rates AND graphical evaluation of feature detection / measurement quality. This is blueFish’s target.

MRM Architecture

Most PFAS analyses are targeted MRM (Multiple Reaction Monitoring). Single sample produces ~100 chromatograms (2 per analyte + internal standards). Current lab workflow uses manual spreadsheet consolidation. blueFish implements template-driven MRM batch processing with automated peak integration.


blueFish Product Specification (from Domain Data)

FeatureSourceEvidence
Full EPA 1633A QC (including ion abundance ratios)Domain expertise“Most labs don’t actually perform all QC checks”
LOQ boundary handlingDomain expertiseState labs report below-LOQ as equal to LOQ
Visual QC for feature detectionDomain expertiseProgenesis QI was good at this but is now dead
Low failure rate peak detectionDomain expertiseMZmine, MS-DIAL, XCMS all have unacceptable failure rates
Sovereign mzML ingestionDomain expertiseChromeleon to msConvert to mzML is the standard path
NIST reference validationDomain expertiseUntargeted PFAS reference spectra as validation baseline
MANA SODA benchmarkingDomain expertiseCross-validation consortium datasets for quality review
Template-driven MRM batch processingDomain expertise~100 chromatograms/sample, tedious manual setup

New Spring Validation Targets

TargetSpring Check
EPA 1633A ion abundance ratio QCFull method compliance checks
LOQ boundary validationBelow-LOQ detection and reporting accuracy
NIST reference spectra matchingUntargeted PFAS validation suite
MANA SODA cross-validation parityMulti-tool benchmark comparison
MRM peak integration accuracy100-chromatogram batch processing

These are validation targets the ecosystem could never have generated internally — they come from decades of analytical chemistry domain expertise.


New Public Data Systems

SystemPurposeStatus
NIST PFAS Reference DataUntargeted PFAS validation baselinePlanned
EPA Method 1633AFull QC compliance specificationPlanned
MANA SODA Benchmark DatasetsCross-validation consortiumPlanned

The gen5 Pattern

Domain expertise (decades of analytical chemistry)
    -> Technical data (NIST, EPA 1633A, MANA SODA, MRM architecture)
    -> blueFish product specification
    -> wetSpring Track 2 new validation targets
    -> blueFish implementation (sovereign PFAS ETL)
    -> Architecture and QC compliance review
    -> blueFish available to any PFAS lab (AGPL-3.0)
    -> Domain knowledge becomes embodied in open tooling

Retirement from the university is not retirement from the science. Decades of analytical chemistry expertise — what tools fail, what QC labs skip, where the data quality gaps are — is now the specification for an open tool that outlives any single institution.