
Susan Holmes
· Associate Professor of BiometryVerifiedStanford University · Symbolic Systems
Active 1984–2026
About
Susan Holmes is a Professor of Statistics, Emerita, at Stanford University. She has been working in non-parametric multivariate statistics applied to Biology since 1985. Holmes trained in the French school of Data Analysis in Montpellier and has taught at MIT, Harvard, and was an Associate Professor of Biometry at Cornell before moving to Stanford in 1998. She created the Thinking Matters class: Breaking Codes and Finding Patterns and enjoys working on big messy data sets, primarily from the areas of Immunology, Cancer Biology, and Microbial Ecology. Her theoretical interests include applied probability, Monte Carlo Markov chains (MCMC), Graph Limit Theory, Differential Geometry, and the topology of the space of Phylogenetic Trees. Holmes co-authored the book Modern Statistics for Modern Biology with Wolfgang Huber from EMBL and teaches this material as a crash course (BIOS221) regularly every year. Her current research focus is on improving the statistical analyses and reproducibility of data in perturbation studies of the Human Microbiome. She holds numerous honors and awards, including fellowships at the Fields Institute and the Center for the Advanced Study of the Behavioral Sciences, and has served on various scientific advisory boards and research institutes.
Research topics
- Computer Science
- Medicine
- Biology
- Immunology
- Data Mining
- Internal medicine
- Data science
- Pathology
- Mathematics
- Mathematical analysis
- Psychology
- Pediatrics
- Geometry
- Pure mathematics
- Bioinformatics
- Computational science
- Programming language
- Algorithm
- Environmental health
- Cell biology
- Anesthesia
- Intensive care medicine
- Virology
- Computational biology
Selected publications
Cell Host & Microbe · 2026-03-27 · 2 citations
articleCell Host & Microbe · 2026-03-18 · 3 citations
articleGuidelines for preventing and reporting contamination in low-biomass microbiome studies
Nature Microbiology · 2025-06-20 · 67 citations
reviewOpen accessScience Translational Medicine · 2025-01-29 · 18 citations
articleOpen accessAt this stage in the COVID-19 pandemic, most infections are "breakthrough" infections that occur in individuals with prior severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) exposure. To refine long-term vaccine strategies against emerging variants, we examined both innate and adaptive immunity in breakthrough infections. We performed single-cell transcriptomic, proteomic, and functional profiling of primary and breakthrough infections to compare immune responses from unvaccinated and vaccinated individuals during the SARS-CoV-2 Delta wave. Breakthrough infections were characterized by a less activated transcriptomic profile in monocytes and natural killer cells, with induction of pathways limiting monocyte migratory potential and natural killer cell proliferation. Furthermore, we observed a female-specific increase in transcriptomic and proteomic activation of multiple innate immune cell subsets during breakthrough infections. These insights suggest that prior SARS-CoV-2 vaccination prevents overactivation of innate immune responses during breakthrough infections with discernible sex-specific patterns and underscore the potential of harnessing vaccines in mitigating pathologic immune responses resulting from overactivation.
SSRN Electronic Journal · 2025-01-01 · 2 citations
preprintOpen accessEstimating the size of a set using cascading exclusion
ArXiv.org · 2025-08-07
preprintOpen accessSenior authorLet $S$ be a finite set, and $X_1,\ldots,X_n$ an i.i.d. uniform sample from $S$. To estimate the size $|S|$, without further structure, one can wait for repeats and use the birthday problem. This requires a sample size of the order $|S|^\frac{1}{2}$. On the other hand, if $S=\{1,2,\ldots,|S|\}$, the maximum of the sample blown up by $n/(n-1)$ gives an efficient estimator based on any growing sample size. This paper gives refinements that interpolate between these extremes. A general non-asymptotic theory is developed. This includes estimating the volume of a compact convex set, the unseen species problem, and a host of testing problems that follow from the question `Is this new observation a typical pick from a large prespecified population?' We also treat regression style predictors. A general theorem gives non-parametric finite $n$ error bounds in all cases.
Author Correction: Community-wide hackathons to identify central themes in single-cell multi-omics
UNC Libraries · 2025-05-13
articleOpen accessCommunity-wide hackathons to identify central themes in single-cell multi-omics
UNC Libraries · 2025-05-13
articleOpen accessmedRxiv · 2025-08-21
preprintOpen accessAbstract Bacterial vaginosis (BV) affects >25% of women worldwide and often recurs after standard-of-care metronidazole (MTZ) treatment. LACTIN-V, a live biotherapeutic product (LBP) containing Lactobacillus crispatus strain CTV-05, reduced recurrent BV in a Phase 2b clinical trial, but efficacy was incomplete. We characterized microbiota and immune effects and correlates of treatment success in trial samples. By week 12, L. crispatus -dominant microbiota was achieved in 30% of LBP recipients compared to 9% of placebo (benefit ratio: 3.31; p<0.005). This effect was mostly due to CTV-05, but native L. crispatus strains were also present and increased over time. Inflammatory cytokines decreased in both arms after MTZ, but returned to baseline in placebo recipients. L. crispatus colonization was associated with pre-MTZ microbiota, baseline cytokine profiles, post-MTZ bacterial load, and clinical and behavioral variables. These findings elucidate LBP microbiota effects and identify predictors of treatment success, informing improved intervention strategies to advance women’s health.
NKp30 and NKG2D contribute to natural killer cell-mediated recognition of HIV-infected cells
iScience · 2025-09-13 · 2 citations
articleOpen access. We characterized the patterns of NK cell ligand expression on CD4 T cells at baseline and after infection with a panel of transmitted/founder HIV-1 strains to identify key receptor-ligand pairings. CRISPR editing of CD4 T cells to knock out the NKp30 ligand B7-H6, or the NKG2D ligand MICB reduced NK cell responses to HIV-infected cells in some donors. Blockade of NKp30 or NKG2D on NK cells compromised their specificity of killing HIV-infected cells. Collectively, we identified receptor-ligand pairs including NKp30:B7-H6 and NKG2D:MICB that contribute to NK cell recognition of HIV-infected cells.
Recent grants
EMSW21-VIGRE: Vertical Integration of Mathematics, Statistics and Applied Mathematics.
NSF · $2.7M · 2005–2014
NIH · $149k · 2010
NSF · $300k · 2012–2015
NIH · $6.2M · 2020
NIH · $895k · 2015
Frequent coauthors
- 85 shared
Catherine A. Blish
Chan Zuckerberg Initiative (United States)
- 67 shared
David A. Relman
Stanford University
- 48 shared
Heather B. Jaspan
University of Cape Town
- 48 shared
Persi Diaconis
- 45 shared
Christof Seiler
Maastricht University
- 44 shared
Clive M. Gray
University of Cape Town
- 37 shared
Anna‐Ursula Happel
University of Cape Town
- 37 shared
Jonathan M. Blackburn
University of Cape Town
Labs
Susan HolmesPI
Awards & honors
- CASBS Fellow, Center for the Advanced study of the Behaviora…
- Breiman Lecturer, N(eur)IPS (December, 2016)
- Fellow, Fields Institute in Mathematical Sciences, Toronto,…
- Director's Transformative Research Award, NIH (2013)
- John Henry Samter University Fellow in Undergraduate Educati…
- Resume-aware match score
- Save to shortlist
- AI-drafted outreach
See your match with Susan Holmes
PhdFit ranks faculty by your research interests, methods, and publications — grounded in their actual work, not templates.
- Free to start
- No credit card
- 30-second signup