Research constellation6 research areas around Giulia C. Kennedy: Gene regulation and gene discovery (UCSF, Millennium, Chiron, Diabetes 1995, Nature Genetics 1995, Proceedings of the National Academy of Sciences of the United States of America 1992); Whole-genome genotyping platforms (Affymetrix, Affymetrix GeneChip Mapping arrays, Whole-genome sampling analysis (one-primer genotyping assay), Bioinformatics 2005, Genome Research 2004, Bioinformatics 2003, Nature Biotechnology 2003); Thyroid nodule genomic classification (Veracyte, Afirma Gene Expression Classifier, Afirma Genomic Sequencing Classifier, Whole-transcriptome RNA sequencing, Ensemble machine learning classifiers, Thyroid 2022, The Journal of Clinical Endocrinology & Metabolism 2021, Frontiers in Endocrinology 2019, Frontiers in Endocrinology 2019, BMC Systems Biology 2019, JAMA Surgery 2018, BMC Bioinformatics 2016, The Journal of Clinical Endocrinology & Metabolism 2012, New England Journal of Medicine 2012, The Journal of Clinical Endocrinology & Metabolism 2010); Machine learning for clinical diagnostics (Veracyte, PinkDx, Ensemble machine learning classifiers, Whole-transcriptome RNA sequencing, International Journal of Gynecological Cancer 2026, Chest 2024, BMC Medical Genomics 2020, BMC Systems Biology 2019, BMC Genomics 2018, JAMA Surgery 2018, Annals of the American Thoracic Society 2017, The Lancet Respiratory Medicine 2015, The Journal of Clinical Endocrinology & Metabolism 2010); Pulmonary genomic classifiers (Veracyte, Percepta Bronchial Genomic Classifier, Envisia Genomic Classifier, Whole-transcriptome RNA sequencing, Ensemble machine learning classifiers, Chest 2024, American Journal of Respiratory and Critical Care Medicine 2021, BMC Medical Genomics 2020, The Lancet Respiratory Medicine 2019, BMC Genomics 2018, Annals of the American Thoracic Society 2017, BMC Cancer 2016, The Lancet Respiratory Medicine 2015); Gynecologic cancer detection (PinkDx, PinkDx vaginal swab classifier for endometrial cancer, Whole-transcriptome RNA sequencing, Ensemble machine learning classifiers, International Journal of Gynecological Cancer 2026).UCSFMillenniumChironDiabetes 1995Nature Genetics 1995PNASUSA 1992AffymetrixAffymetrix GeneChip Mappi…Whole-genome sampling ana…Bioinformatics 2005Genome Research 2004Bioinformatics 2003Nature Biotechnology 2003VeracyteAfirma Gene Expression Cl…Afirma Genomic Sequencing…Whole-transcriptome RNA s…Ensemble machine learning…Thyroid 2022JCEM 2021Frontiers in Endocrinology 2019Frontiers in Endocrinology 2019BMC Systems Biology 2019JAMA Surgery 2018BMC Bioinformatics 2016JCEM 2012New England Journal of Medicine 2012JCEM 2010PinkDxIJGC 2026Chest 2024BMC Medical Genomics 2020BMC Genomics 2018AATS 2017Lancet Respiratory Medicine 2015Percepta Bronchial Genomi…Envisia Genomic ClassifierAJRCCM 2021Lancet Respiratory Medicine 2019BMC Cancer 2016PinkDx vaginal swab class…Gene regulation and gene discovery1992 – 1999Whole-genome genotyping platformsJan 2000 – Mar 2008Thyroid nodule genomic classificationApr 2008 – Dec 2022Machine learning for clinical diagnosticsDec 2010 – presentPulmonary genomic classifiersApr 2015 – Apr 2024Gynecologic cancer detection2022 – present
Research areas and the organisations, technologies and publications connected to them
Research areaPeriodConnected organisations, technologies and publications
Gene regulation and gene discovery1992 – 1999UCSF · Millennium · Chiron · Diabetes 1995 · Nature Genetics 1995 · Proceedings of the National Academy of Sciences of the United States of America 1992
Whole-genome genotyping platformsJan 2000 – Mar 2008Affymetrix · Affymetrix GeneChip Mapping arrays · Whole-genome sampling analysis (one-primer genotyping assay) · Bioinformatics 2005 · Genome Research 2004 · Bioinformatics 2003 · Nature Biotechnology 2003
Thyroid nodule genomic classificationApr 2008 – Dec 2022Veracyte · Afirma Gene Expression Classifier · Afirma Genomic Sequencing Classifier · Whole-transcriptome RNA sequencing · Ensemble machine learning classifiers · Thyroid 2022 · The Journal of Clinical Endocrinology & Metabolism 2021 · Frontiers in Endocrinology 2019 · Frontiers in Endocrinology 2019 · BMC Systems Biology 2019 · JAMA Surgery 2018 · BMC Bioinformatics 2016 · The Journal of Clinical Endocrinology & Metabolism 2012 · New England Journal of Medicine 2012 · The Journal of Clinical Endocrinology & Metabolism 2010
Machine learning for clinical diagnosticsDec 2010 – presentVeracyte · PinkDx · Ensemble machine learning classifiers · Whole-transcriptome RNA sequencing · International Journal of Gynecological Cancer 2026 · Chest 2024 · BMC Medical Genomics 2020 · BMC Systems Biology 2019 · BMC Genomics 2018 · JAMA Surgery 2018 · Annals of the American Thoracic Society 2017 · The Lancet Respiratory Medicine 2015 · The Journal of Clinical Endocrinology & Metabolism 2010
Pulmonary genomic classifiersApr 2015 – Apr 2024Veracyte · Percepta Bronchial Genomic Classifier · Envisia Genomic Classifier · Whole-transcriptome RNA sequencing · Ensemble machine learning classifiers · Chest 2024 · American Journal of Respiratory and Critical Care Medicine 2021 · BMC Medical Genomics 2020 · The Lancet Respiratory Medicine 2019 · BMC Genomics 2018 · Annals of the American Thoracic Society 2017 · BMC Cancer 2016 · The Lancet Respiratory Medicine 2015
Gynecologic cancer detection2022 – presentPinkDx · PinkDx vaginal swab classifier for endometrial cancer · Whole-transcriptome RNA sequencing · Ensemble machine learning classifiers · International Journal of Gynecological Cancer 2026
Research area 01

Gene regulation and gene discovery

1992 – 1999

Kennedy's first research programme ran from insulin gene transcription to disease gene discovery. As a postdoctoral scientist in William J. Rutter's laboratory at UCSF she identified Pur-1, a zinc-finger transactivator of the insulin promoter (PNAS, 1992), and showed that the minisatellite at the type 1 diabetes locus IDDM2 regulates insulin transcription (Nature Genetics, 1995). At Millennium Pharmaceuticals she implemented genomic and genetic approaches to uncover diabetes susceptibility genes, and at Chiron Corporation she led colon and breast cancer gene discovery efforts that identified oncology markers for therapeutic drug development.123456

  • Insulin gene transcription
  • Type 1 diabetes genetics
  • Diabetes susceptibility genes
  • Colon and breast cancer markers
Research area 02

Whole-genome genotyping platforms

JAN 2000 – MAR 2008

At Affymetrix, where she was a Senior Director from January 2000 to March 2008 leading the Genomics Collaborations and Genotyping Technology R&D groups, Kennedy's teams published the methods that made array-based SNP genotyping practical: "Large-scale genotyping of complex DNA" (Nature Biotechnology, 2003, first author), the one-primer 10K assay (Genome Research, 2004), the 100K array set (Nature Methods, 2004) and the genotype-calling algorithms in Bioinformatics (2003, 2005). She is a co-inventor on the Affymetrix patent "Method for genotyping polymorphisms in humans".547891011

  • SNP genotyping
  • Genome-wide association studies
  • HapMap-era human genetics
  • Copy-number detection

Organisations

Technologies

  • Affymetrix GeneChip Mapping arrays · PLATFORM

    High-density oligonucleotide arrays for genotyping single-nucleotide polymorphisms genome-wide. The 10K array genotyped over 10,000 SNPs per individual on a single array (Genome Research, 2004) and the 100K set scored 116,204 SNPs on a pair of arrays with call rates above 99% (Nature Methods, 2004). A 2006 NHGRI primer cites "Kennedy et al, Nature Biotech, 2003" as the reference for the GeneChip Mapping Assay, and Kennedy is a co-inventor on Affymetrix's US 7,300,788, "Method for genotyping polymorphisms in humans".

  • Whole-genome sampling analysis (one-primer genotyping assay) · METHOD

    The sample-preparation method behind the GeneChip Mapping arrays: restriction digestion fractionates the genome, a specific fractionated subset is amplified with a single primer, and alleles are read by allele-specific hybridization to the array. Described in "Parallel Genotyping of Over 10,000 SNPs Using a One-Primer Assay on a High-Density Oligonucleotide Array" (Genome Research, 2004), on which Kennedy is a co-author, and introduced in "Large-scale genotyping of complex DNA" (Nature Biotechnology, 2003), on which she is first author.

Chapters

Research area 03

Thyroid nodule genomic classification

APR 2008 – DEC 2022

The programme that defined Veracyte, where Kennedy was Chief Scientific Officer from 2008 and Chief Medical Officer from 2018. The Afirma Gene Expression Classifier was developed from 315 nodules (JCEM, 2010), launched in January 2011 and validated prospectively in 3,789 patients (NEJM, 2012). The RNA sequencing-based Afirma GSC followed in 2017, validated in JAMA Surgery (2018) with at least one third more benign calls than the original test, and was extended with the Xpression Atlas of variants and fusions, Hürthle cell and medullary thyroid carcinoma classifiers. By January 2018 Veracyte had performed 100,000 Afirma tests.5121314151617

  • Indeterminate thyroid fine-needle aspirates
  • Avoiding diagnostic thyroid surgery
  • Hürthle cell nodules
  • Medullary thyroid carcinoma
  • Variant and fusion reporting

Organisations

Technologies

  • Afirma Gene Expression Classifier · PRODUCT

    Veracyte's first commercial product, launched in January 2011 as the Afirma Thyroid FNA Analysis. It combines cytopathology with a gene expression test, described in trade press as a 142-gene signature, to identify benign nodules among thyroid fine-needle aspirates that cytology leaves indeterminate. Developed from 315 nodules and more than 247,186 transcripts (JCEM, 2010) and validated prospectively at 49 sites in 3,789 patients (NEJM, 2012), with Kennedy as last and second author respectively.

  • Afirma Genomic Sequencing Classifier · PRODUCT

    The second-generation Afirma test, built on whole-transcriptome RNA sequencing and ensemble machine learning algorithms. Veracyte announced pivotal validation data on May 4, 2017 and began transitioning patients to the GSC that quarter, saying it could save an estimated 70% of patients with benign nodules from unnecessary surgery. The JAMA Surgery validation (2018) found it identifies at least one third more benign nodules than the original test with 91% sensitivity. The companion Afirma Xpression Atlas, launched May 2018, reports gene variants and fusions from the same sample.

  • Whole-transcriptome RNA sequencing · PLATFORM

    The sequencing platform that replaced microarrays in Veracyte's second-generation tests. Kennedy described it at the 2018 Xpression Atlas launch: "Our RNA sequencing-based platform enables us to derive unprecedented amounts of rich genomic content from clinical samples." The Afirma GSC, Percepta GSC and Envisia classifiers are built on it, and PinkDx applies whole-transcriptome sequencing to vaginal swab samples in its endometrial cancer programme.

  • Ensemble machine learning classifiers · METHOD

    The algorithmic approach behind Veracyte's sequencing-era classifiers and PinkDx's swab test. Kennedy said in 2017 that "the ensemble machine learning algorithms, which underlie the performance of Afirma GSC, derive from years of thoughtful scientific research, data analysis and statistical modeling", and that the company was "employing the same machine learning methods that are being used in other fields such as social media and self-driving cars". The 2019 BMC Systems Biology paper, with Kennedy as last author, describes a trio of machine learning algorithms for Hürthle cell nodules.

Publications

Chapters

Research area 04

Machine learning for clinical diagnostics

DEC 2010 – PRESENT

The method that links every classifier Kennedy has helped build since 2010: training algorithms on high-dimensional genomic data from clinical samples and validating them against histopathology. She presented on Veracyte's use of machine learning to translate genomic data into commercial diagnostic tests at the Precision Medicine World Conference in January 2018, and described the Afirma GSC's ensemble machine learning algorithms as deriving "from years of thoughtful scientific research, data analysis and statistical modeling". The same approach underlies the Envisia and Percepta classifiers and PinkDx's whole-transcriptome vaginal swab test.1815192021

  • Genomic classifier development
  • Analytical and clinical validation
  • Ensemble algorithms
  • Translating sequencing data into diagnostic tests

Organisations

Technologies

  • Ensemble machine learning classifiers · METHOD

    The algorithmic approach behind Veracyte's sequencing-era classifiers and PinkDx's swab test. Kennedy said in 2017 that "the ensemble machine learning algorithms, which underlie the performance of Afirma GSC, derive from years of thoughtful scientific research, data analysis and statistical modeling", and that the company was "employing the same machine learning methods that are being used in other fields such as social media and self-driving cars". The 2019 BMC Systems Biology paper, with Kennedy as last author, describes a trio of machine learning algorithms for Hürthle cell nodules.

  • Whole-transcriptome RNA sequencing · PLATFORM

    The sequencing platform that replaced microarrays in Veracyte's second-generation tests. Kennedy described it at the 2018 Xpression Atlas launch: "Our RNA sequencing-based platform enables us to derive unprecedented amounts of rich genomic content from clinical samples." The Afirma GSC, Percepta GSC and Envisia classifiers are built on it, and PinkDx applies whole-transcriptome sequencing to vaginal swab samples in its endometrial cancer programme.

Chapters

Research area 05

Pulmonary genomic classifiers

APR 2015 – APR 2024

Veracyte's second and third commercial tests took the thyroid approach to the lung. Percepta, launched April 16, 2015, classifies lung nodules that are inconclusive by bronchoscopy from bronchial brushings; Envisia, launched October 25, 2016, identifies the usual interstitial pneumonia pattern of idiopathic pulmonary fibrosis in transbronchial biopsies without surgical lung biopsy, and received Medicare coverage effective April 1, 2019. Kennedy is an author of the Envisia development and validation papers (Annals ATS 2017, BMC Genomics 2018, Lancet Respiratory Medicine 2019, AJRCCM 2021), the Percepta analytical and GSC papers, and the 2024 Chest validation of the Percepta Nasal Swab.2223242526274

  • Lung nodules after inconclusive bronchoscopy
  • Idiopathic pulmonary fibrosis diagnosis
  • Usual interstitial pneumonia pattern
  • Nasal swab lung cancer risk assessment

Organisations

Technologies

  • Percepta Bronchial Genomic Classifier · PRODUCT

    Veracyte's lung cancer diagnostic, launched April 16, 2015, which classifies lung nodules and lesions that are inconclusive by bronchoscopy using bronchial brushing specimens. The next-generation Percepta Genomic Sequencing Classifier became available to physicians on June 26, 2019, and on May 19, 2021 Veracyte announced pivotal validation data for the Percepta Nasal Swab test for lung cancer risk, with Kennedy quoted as chief scientific officer and chief medical officer.

  • Envisia Genomic Classifier · PRODUCT

    Veracyte's classifier for idiopathic pulmonary fibrosis, launched at the CHEST Annual Meeting on October 25, 2016, where Kennedy presented clinical validity data. It identifies the usual interstitial pneumonia pattern in conventional transbronchial lung biopsy samples, offering an alternative to surgical lung biopsy; the prospective validation appeared in The Lancet Respiratory Medicine in 2019 and a study of the classifier as a complement to high-resolution CT in AJRCCM in 2021. Medicare coverage through the Palmetto GBA MolDX program took effect April 1, 2019.

  • Whole-transcriptome RNA sequencing · PLATFORM

    The sequencing platform that replaced microarrays in Veracyte's second-generation tests. Kennedy described it at the 2018 Xpression Atlas launch: "Our RNA sequencing-based platform enables us to derive unprecedented amounts of rich genomic content from clinical samples." The Afirma GSC, Percepta GSC and Envisia classifiers are built on it, and PinkDx applies whole-transcriptome sequencing to vaginal swab samples in its endometrial cancer programme.

  • Ensemble machine learning classifiers · METHOD

    The algorithmic approach behind Veracyte's sequencing-era classifiers and PinkDx's swab test. Kennedy said in 2017 that "the ensemble machine learning algorithms, which underlie the performance of Afirma GSC, derive from years of thoughtful scientific research, data analysis and statistical modeling", and that the company was "employing the same machine learning methods that are being used in other fields such as social media and self-driving cars". The 2019 BMC Systems Biology paper, with Kennedy as last author, describes a trio of machine learning algorithms for Hürthle cell nodules.

Chapters

Research area 06

Gynecologic cancer detection

2022 – PRESENT

At PinkDx, co-founded with Bonnie Anderson and launched publicly on April 25, 2024 with a $40 million Series A, Kennedy leads scientific strategy and translational research applying whole-transcriptome sequencing and machine learning to gynecologic cancer, starting with endometrial cancer. The exploratory PNK001 study enrolled 236 women undergoing hysterectomy in 2024 and showed that molecular signals of endometrial cancer can be detected from a vaginal swab (International Journal of Gynecological Cancer, April 2026); the PROACTION study (NCT06527157) is validating the test across 16 U.S. sites.28293031213233

  • Endometrial cancer detection from vaginal swabs
  • Less invasive diagnostics for abnormal uterine bleeding
  • Gynecologic cancer diagnostics

Organisations

Technologies

  • PinkDx vaginal swab classifier for endometrial cancer · ASSAY

    PinkDx's first product in development: a clinician-collected vaginal swab, with no speculum required, for endometrial cancer in women with abnormal bleeding. The sample is shipped to the PinkDx laboratory in California and an algorithm classifies it as positive or not positive for endometrial cancer. The exploratory PNK001 study (NCT06460818) showed that molecular signals of endometrial cancer can be read from a vaginal swab by whole-transcriptome sequencing and machine learning (IJGC, 2026), and the PROACTION study (NCT06527157) is validating the test.

  • Whole-transcriptome RNA sequencing · PLATFORM

    The sequencing platform that replaced microarrays in Veracyte's second-generation tests. Kennedy described it at the 2018 Xpression Atlas launch: "Our RNA sequencing-based platform enables us to derive unprecedented amounts of rich genomic content from clinical samples." The Afirma GSC, Percepta GSC and Envisia classifiers are built on it, and PinkDx applies whole-transcriptome sequencing to vaginal swab samples in its endometrial cancer programme.

  • Ensemble machine learning classifiers · METHOD

    The algorithmic approach behind Veracyte's sequencing-era classifiers and PinkDx's swab test. Kennedy said in 2017 that "the ensemble machine learning algorithms, which underlie the performance of Afirma GSC, derive from years of thoughtful scientific research, data analysis and statistical modeling", and that the company was "employing the same machine learning methods that are being used in other fields such as social media and self-driving cars". The 2019 BMC Systems Biology paper, with Kennedy as last author, describes a trio of machine learning algorithms for Hürthle cell nodules.

Chapters