Faculty Spring Profiles

Status: Working draft
Purpose: Map known faculty to ecoPrimals springs, identify candidate papers for reproduction
Last Updated: February 26, 2026

How This Document Works

Each professor is profiled with:

1. Connection — how we know them and what spring(s) their work maps to
2. Key Papers — candidate publications for Phase A reproduction in the springs
3. BarraCuda Relevance — which GPU primitives their work exercises
4. Status — what has already been reproduced vs. what is candidate work

hotSpring — Computational Plasma Physics

Michael Murillo

Associate Professor, CMSE, MSU
https://engineering.msu.edu/faculty/michael-murillo
Connection: MSDS professor (master’s program)

Spring Status: ALL PHASES COMPLETE — 22 papers, ~700 checks, 39/39 suites. Exp 022 (live NPU, 32⁴ production) finished Feb 27.

Coffee meeting scheduled: Tuesday March 3, 2026. Technical summary prepared: hotSpring validation summary.

Reproduced Papers (Murillo Group):

Candidate Papers for Future Reproduction (Tier 4 — WDM):

• Diaw & Murillo (2023) “Generalized Hydrodynamics Model for Strongly Coupled Plasmas”
• Murillo (2025) “Computational barriers” (arXiv:2505.02494) — WDM roadmap paper

BarraCuda primitives exercised: GEMM, Velocity Verlet, Yukawa/Coulomb force kernels, MLP surrogate, RBF interpolation, Green-Kubo transport, Sturm bisection, DF64 arithmetic, SU(3) lattice gauge, HMC, ESN reservoir, NPU streaming

airSpring — Evapotranspiration & Precision Irrigation

Younsuk Dong

Assistant Professor, Biosystems & Agricultural Engineering, MSU
https://www.egr.msu.edu/bae/water/irrigation/
Connection: New lab job (2026)

Spring Status: 3,123+ total checks (594 Python + 491 Rust + 570 validation + 1393 atlas + 75 cross-val), 22 experiments, 27 binaries

Reproduced Papers:

Experiments (beyond original 3 Dong papers): Richards PDE, biochar P adsorption, dual Kc, cover crops, yield response, scheduling optimization, lysimeter, sensitivity analysis, atlas, PT/HG/Thornthwaite ET₀, GDD, Saxton-Rawls pedotransfer, CW2D, 60-year water balance.

Candidate Papers for Future Reproduction:

• Dong et al. — IoT soil moisture sensor network calibration (fieldwork data)
• Allen et al. (1998) FAO-56 — extended crop coefficient studies
• Regional ET₀ model comparisons across Michigan microclimates

BarraCuda primitives exercised: MLP surrogate (FAO-56 approximation), time-series LSTM (weather forecasting), data pipeline batch processing

wetSpring — Microbial Ecology, Phage Biology, Environmental Chemistry

Jesse Cahill

Senior MTS, Sandia National Laboratories (Bioscience)
Connection: Time at Sandia

Spring Area: Track 1 — Life Science (algae ponds, phage biocontrol)

Candidate Papers for Reproduction:

• Cahill et al. — Phage-mediated biocontrol in algal raceway ponds
• Phage lifecycle dynamics and predator-prey oscillations in bioreactor systems
• Algal pond crash forensics — temporal metagenomics

BarraCuda relevance: Time-series anomaly detection (pond crash prediction), population dynamics ODE solvers

Chuck Smallwood

Principal MTS, Sandia National Laboratories (Bioscience)
Connection: Time at Sandia

Spring Area: Track 1 — Life Science (metagenomics, microbial community monitoring)

Candidate Papers for Reproduction:

• Smallwood et al. — Raceway pond metagenomic surveillance pipelines
• Microbial community stability metrics under perturbation

BarraCuda relevance: Sequence alignment (GEMM-heavy), dimensionality reduction, diversity index computation

A. Daniel Jones

Professor, Biochemistry & Molecular Biology / Chemistry, MSU
https://www.canr.msu.edu/news/center-for-pfas-research-faculty-spotlight-a-daniel-jones
Connection: PFAS job

Spring Area: Track 2 — PFAS / blueFish

Candidate Papers for Reproduction:

• Jones et al. — PFAS mass spectrometry detection pipelines
• High-resolution mass spec peak identification and quantification
• Environmental PFAS fate-and-transport modeling

BarraCuda relevance: Signal processing (FFT, peak detection), spectral analysis shaders, anomaly detection in analytical chemistry data

Christopher Waters

Professor, Microbiology, Genetics & Immunology, MSU
https://directory.natsci.msu.edu/directory/Profiles/Person/101708
https://mgi.natsci.msu.edu/labs/waters-lab/
watersc3@msu.edu
Connection: Undergraduate professor

Spring Area: wetSpring Track 1 — Microbial signaling, biofilm dynamics, quorum sensing

Key Research Themes:

1. c-di-GMP signaling — second messenger controlling biofilm ↔ motility switch in V. cholerae
2. Quorum sensing — density-dependent gene regulation via autoinducers
3. Integration of signaling pathways — c-di-GMP + quorum sensing convergence
4. Phage defense — deoxycytidine deaminase protection (Nature Microbiology 2022)
5. Cancer immunotherapy — cyclic di-nucleotides as immune adjuvants

Candidate Papers for Reproduction:

BarraCuda relevance: ODE/PDE solvers (reaction-diffusion for signaling), stochastic simulation (Gillespie algorithm for quorum sensing), bifurcation analysis, graph computation (regulatory networks → neuralAPI pathway graphs)

Master’s Program Professors — Computational Methods

Emily Dolson

Assistant Professor, Computer Science & Engineering, MSU
https://engineering.msu.edu/directory/faculty/dolsonem
dolsonem@msu.edu
Connection: Master’s program professor
Core faculty: Ecology, Evolution, and Behavior program

Key Research Themes:

1. Controlling evolutionary trajectories — counterdiabatic driving applied to evolution
2. Open-ended evolution — measuring when systems produce genuine novelty
3. Eco-evolutionary dynamics — ecology and evolution as coupled processes
4. Mathematical oncology — evolutionary dynamics of cancer

Candidate Papers for Reproduction:

BarraCuda relevance: Fitness landscape evaluation (GEMM), population simulation (parallel agents), evolutionary optimization (genetic algorithm shaders), diversity metrics computation

Critical connection to Constrained Evolution Formal: Dolson’s work on counterdiabatic driving of evolution is the closest published analog to the constrained evolution methodology described in Constrained Evolution Formal. Reproducing Iram et al. (2020) would provide external validation of the theoretical framework.

Kevin Liu

Associate Professor, CMSE, MSU
https://engineering.msu.edu/directory/faculty/kjl
kjl@msu.edu
Connection: Master’s program professor
Faculty in: Genetics & Genome Sciences, Ecology, Evolution & Behavior

Key Research Themes:

1. Phylogenetic inference at scale — SATé/SATé-II for massive tree estimation
2. Introgression detection — PhyloNet-HMM for gene flow across species
3. Comparative genomics — detecting adaptive introgression in eukaryotes
4. Resampling methods — bootstrap and RAWR for phylogenetic confidence

Candidate Papers for Reproduction:

BarraCuda relevance: Sequence alignment (GEMM-heavy Smith-Waterman), HMM forward/backward/Viterbi (matrix operations), phylogenetic likelihood computation (parallel tree evaluation), bootstrap resampling (embarrassingly parallel)

Working manuscripts to watch:

• “A phylogenomic study of adaptive co-evolution between early diverging fungi and obligate bacterial endosymbionts” — direct wetSpring relevance
• “Scalable statistical inference of species phylogenies from large-scale resequenced genomic datasets” — HPC/GPU candidate

Alexei Bazavov

Associate Professor, CMSE & Physics & Astronomy, MSU
https://directory.natsci.msu.edu/directory/Profiles/Person/101033
bazavov@msu.edu
Connection: Master’s program professor
Affiliations: CERN Theory Division, Fermilab Lattice, HPQCD, MILC Collaborations

Key Research Themes:

1. Lattice QCD — ab initio computation of strong force properties
2. Hadronic vacuum polarization — precision calculation for muon g-2
3. Inverse problems — spectral reconstruction from lattice data
4. Parallel algorithms — molecular dynamics on lattice gauge configurations

Reproduced Papers:

Production Results: Exp 013 (32⁴, native f64, 13.6h, β_c=5.69) and Exp 022 (32⁴, DF64+NPU, 14.2h, 10 NPU-steered β points, 5,900 measurements). Deconfinement transition confirmed. DF64 discovery: 9.9× native f64 throughput.

Candidate Papers for Future Reproduction:

BarraCuda primitives exercised: SU(3) GEMM (link multiplication), HMC molecular dynamics, Dirac CG solver, plaquette/Polyakov/susceptibility observables, DF64 arithmetic, PRNG (Philox), WGSL shaders (25).

Critical connection to hotSpring: Bazavov’s lattice QCD and Murillo’s plasma physics are both studying strongly coupled many-body systems. The computational methods overlap significantly — both use molecular dynamics, both need equation of state calculations, both deal with long-range correlations. The shared BarraCuda kernel library now serves both.

Ilya Kachkovskiy

Assistant Professor, Department of Mathematics, MSU
https://users.math.msu.edu/users/ikachkov/
ikachkov@msu.edu
Connection: Sold the NucBox M6 on Facebook Marketplace; brief GPU conversation
Previously at: Institute for Advanced Study, UC Irvine
NSF funded: DMS-1758326 “Spectral theory of periodic and quasiperiodic quantum systems”

Key Research Themes:

1. Anderson localization — how disorder causes quantum waves to localize (absence of transport in disordered media)
2. Spectral theory of quasiperiodic operators — eigenvalue structure of systems that are “almost periodic” (structured noise)
3. Transport in quantum spin systems — energy/information propagation through spin chains
4. Almost commuting operators — approximate symmetries in quantum systems (C*-algebra framework)

Reproduced Papers:

Candidate Papers for Future Reproduction:

BarraCuda relevance: Eigenvalue solvers (Lanczos — validated on GPU), sparse matrix-vector products, Hofstadter butterfly computation, Anderson localization IPR. The spectral methods require f64 precision — same requirement as hotSpring’s plasma MD and Bazavov’s lattice QCD.

Why this matters: Kachkovskiy provides the mathematical layer that sits between Murillo’s physics (transport in classical plasmas) and Bazavov’s physics (transport in quantum field theories). His spectral theory is the eigenvalue mathematics both of them use but neither of them proves. Anderson localization — the core of his research — is the rigorous mathematical framework for “when does signal propagate vs. when does noise trap it?” That’s groundSpring’s central question stated in the language of quantum mechanics.

Co-author network: Jean Bourgain (Fields Medalist, IAS — deceased 2018), Svetlana Jitomirskaya (UCI, Dannie Heineman Prize), Nikolay Filonov (Steklov Institute), Yuri Safarov (King’s College London). This is a tier-1 mathematics pedigree.

Rika Anderson

Associate Professor, Department of Biology, Carleton College
https://www.carleton.edu/directory/randerson/
randerson@carleton.edu
Connection: Found via literature search — her work on Sulfolobus in Yellowstone hot springs is the living experimental corollary to the Taq polymerase argument in Constrained Evolution Formal
Previously at: University of Washington (MS, PhD); Virtual Planetary Laboratory (NASA)

Key Research Themes:

1. Microbial evolution in deep-sea hydrothermal vents — how extreme environments shape microbial genomes (Nature Communications 2017)
2. Stochastic vs deterministic evolution in low-biomass environments — when does genetic drift dominate natural selection? (mSystems 2021, 2022)
3. Population genomics of extremophiles — Sulfolobus in Yellowstone, Sulfurovum at vents, subseafloor archaea
4. Viral ecology and host-virus coevolution — CRISPRs as metagenomic tools, phage biogeography
5. Pangenomics and gene gain/loss — selection vs drift in functional evolution

Candidate Papers for Reproduction:

BarraCuda relevance: Sequence alignment (GEMM), diversity indices (reduction), dN/dS selection tests (pairwise comparison), phylogenetic tree construction (parallel likelihood), rarefaction curves (bootstrap resampling), pangenome analysis (set operations). All of these are already validated in wetSpring’s sovereign pipeline.

Why this is the corollary to Taq polymerase: The constrained evolution thesis (§1.1) uses Thermus aquaticus from Yellowstone hot springs as its founding biological example — thermal constraint produced Taq polymerase, which enabled PCR and modern molecular biology. Anderson’s lab has published the population genomics of another extremophile (Sulfolobus islandicus) living in the exact same Yellowstone hot springs. Her 2017 paper with Campbell shows how thermal constraint drives population differentiation, susceptibility to mobile genetic elements, and structured genomic variation in these hot spring populations. This is not a metaphor for the constrained evolution thesis — this is the empirical data that would appear in §1.1 if we were writing a full literature review. Additionally, her 2021 mSystems paper explicitly discusses stochastic vs deterministic evolution under environmental constraint, cites Lenski’s LTEE (the same experiment that anchors §1.2), and introduces Muller’s ratchet as a consequence of extreme energy limitation — all themes directly present in Constrained Evolution Formal.

Co-author network: John Baross (UW — pioneer of deep-sea microbiology), Julie Huber (WHOI — subseafloor biosphere), Mitch Sogin (MBL — rare biosphere), Rachel Whitaker (Illinois — Sulfolobus population genetics), Emily Stüeken (St Andrews — early Earth geochemistry), Ben Tully (USC — marine metagenomics).

MSU Drug Discovery Program — Pharmacology & Toxicology

Erika Lisabeth

Director, Assay Development and Drug Repurposing Core (ADDRC), Pharmacology & Toxicology, MSU https://drugdiscovery.msu.edu/about-us/people/erika-lisabeth-ph-d.aspx Connection: Referred by Gonzales (March 2026 interview)

Key Research Themes:

1. HTS assay development — converting bench-top assays to high-throughput screening format
2. Drug repurposing — identifying new indications for existing compounds
3. EphA3 receptor tyrosine kinase — somatic mutations inactivate EphA3 in cancer (postdoc work)
4. NF-κB degradation pathway — characterized inhibitor degradation (UCSD PhD)

Candidate Papers for Reproduction:

ADDRC Infrastructure:

• 8,000+ compound library
• Liquid-handling robots, automated plate readers, high-content microscopes
• GREENScreen informatics for compound management
• Available 24/7 to MSU researchers and external biotech

Spring Relevance: The ADDRC is the institutional screening platform for the Anderson-augmented MATRIX drug repurposing scores (nS-605). Pipeline: computational scoring → ADDRC HTS → Gonzales iPSC validation. The Brucella screen (8,000 compounds) demonstrates the throughput; applying Anderson geometry scoring to compound selection is the extension.

BarraCuda relevance: Plate reader data processing (batch reduction), dose-response curve fitting (Hill equation — already in nS-601), diversity metrics for compound clustering, MATRIX scoring automation

Richard Neubig

Professor Emeritus, Director Drug Discovery Program, Pharmacology & Toxicology, MSU https://phmtox.msu.edu/people/rneubig Connection: MSU Drug Discovery leadership; referred via Gonzales (March 2026)

Key Research Themes:

1. GPCR signaling — G-protein coupled receptor pharmacology (>25% of current drug targets)
2. Rho/MRTF/SRF gene transcription — small molecule inhibitors for fibrotic diseases
3. Skin fibrosis — Rho pathway inhibitors in dermal fibrosis models
4. Melanoma metastasis — MRTF/SRF inhibitors block metastatic phenotype
5. Academic drug discovery — founded UMich Center for Chemical Genomics, established MSU Drug Discovery

Candidate Papers for Reproduction:

Spring Relevance: Neubig’s skin fibrosis work connects directly to Paper 12: Immunological Anderson’s dimensional promotion model — fibrosis changes tissue geometry, which changes Anderson localization of cytokine signals. If Rho/MRTF/SRF cross-talks with JAK/STAT in skin, his inhibitors become candidates for Anderson-augmented MATRIX scoring.

BarraCuda relevance: GPCR docking (future — structural), dose-response fitting (immediate — Hill equation), signaling pathway graph analysis

Edmund Ellsworth

Interim Director Drug Discovery, Director Medicinal Chemistry, MSU https://drugdiscovery.msu.edu/about-us/people/index.aspx Connection: MSU Drug Discovery leadership

Spring Relevance: Downstream of HTS — medicinal chemistry optimization after ADDRC screening identifies hits. Not a direct spring reproduction target yet, but the endpoint of the computation → screening → chemistry pipeline.

Cross-Spring Faculty Connections

                    hotSpring (Murillo, Bazavov, Kachkovskiy, R. Anderson)
                    ┌──────────────────────────┐
                    │ Strongly coupled           │
                    │ many-body systems          │
                    │ MD, EOS, spectral theory   │
                    │ GPU compute                │
                    │ Hot spring microbial pop.   │
                    │ genetics (Taq corollary)   │
                    └──────┬─────────────────────┘
                           │
          ┌────────────────┼────────────────┐
          │                │                │
   groundSpring            │         neuralSpring
   (Bazavov inverse,       │    (Dolson evolution,
    Kachkovskiy localize,  │     Liu HMM/phylo,
    R. Anderson stochastic)│     Bazavov parallel,
   ┌──────────────┐        │     Kachkovskiy spectral,
   │ Inverse probs│        │     R. Anderson pangenomics,
   │ Noise/signal │        │     Gonzales dose-response)
   │ Anderson loc │        │    ┌──────────────┐
   │ Stochastic vs│        │    │ ML primitives │
   │ deterministic│        │    │ Evolutionary  │
   │ Spectral recon│       │    │ optimization  │
   └──────┬───────┘        │    └──────┬───────┘
          │                │           │
          └────────┬───────┘───────────┘
                   │
            wetSpring (Cahill, Smallwood, Jones, Waters, Liu, R. Anderson,
                      Gonzales, Lisabeth, Neubig)
            ┌──────────────────────────┐
            │ Microbial ecology         │
            │ Metagenomics              │
            │ Quorum sensing/signaling  │
            │ PFAS detection            │
            │ Phage dynamics            │
            │ Vent population genomics  │
            │ Deep-time enzyme evolution│
            │ Immunological Anderson    │
            │ Drug repurposing (MATRIX) │
            │ HTS / ADDRC screening     │
            └──────────────────────────┘

Priority Reproduction Candidates (Next Phase)

Ranked by: (1) direct spring relevance, (2) reproducibility, (3) BarraCuda kernel coverage

Tier 1 — High priority, clear reproduction path

1. Iram, Dolson et al. (2020) — Nature Physics — Counterdiabatic evolution control → validates constrained evolution thesis
2. Waters et al. (2008) — J Bacteriology — c-di-GMP/QS biofilm model → ODE system, fully specified
3. Liu et al. (2014) — PLoS Comp Bio — PhyloNet-HMM introgression → HMM implementation validates sequence model primitives
4. Bazavov [HotQCD] (2014) — Nuclear Physics A — QCD EOS → extends hotSpring beyond plasma

Tier 2 — Strong candidates, may need data access

5. Dolson et al. (2019) — MODES Toolbox — Open-ended evolution metrics → apply to BarraCuda’s own evolution
6. Massie et al. (2012) — PNAS — c-di-GMP signaling specificity → quantitative signaling model
7. Liu et al. (2009) — Science — SATé phylogenetic estimation → large-scale divide-and-conquer benchmark
8. Hsueh et al. (2022) — Nature Microbiology — Phage defense deaminase → evolutionary arms race dynamics

Tier 2 (new) — Empirical constrained evolution evidence

9. Campbell, Anderson et al. (2017) — Env Microbiol — Sulfolobus population genetics in Yellowstone hot springs → direct empirical data for Constrained Evolution Formal §1.1 (same environment as Taq polymerase)
10. Anderson (2021) — mSystems — Stochastic vs deterministic evolution in subsurface → supports §1.2 Lenski argument, cites LTEE, formalizes when drift dominates selection
11. Moulana, Anderson et al. (2020) — mSystems — Constrained evolution of Sulfurovum pangenomes → gene gain/loss driven by geochemistry = functional evolution under environmental constraint

Tier 2.5 — Mathematical foundations, strengthens multiple springs

12. Bourgain & Kachkovskiy (2018) — GAFA — Anderson localization for interacting particles → groundSpring noise/signal theory, computational eigenvalue methods for BarraCuda
13. Kachkovskiy (2016) — CMP — Quasiperiodic spin chain transport → hotSpring plasma transport bridge, validates Lanczos/spectral BarraCuda primitives

Tier 2.7 — Drug Discovery / Pharmacology (MSU Drug Discovery Program)

14. Lisabeth et al. (2024) — Spartan Med Res J — Brucella HTS screen (8,000+ compounds) → validates ADDRC pipeline throughput, demonstrates computation→screening workflow for Anderson-augmented MATRIX scoring
15. Neubig group — Rho/MRTF/SRF skin fibrosis inhibitors → wetSpring + neuralSpring: If Rho pathway cross-talks with JAK/STAT (Paper 12: Immunological Anderson §8 Q7), Rho inhibitors become Anderson-scorable candidates for AD. Dose-response modeling uses same Hill equation as nS-601
16. Gonzales (2014–2021) — 6 papers (G1–G6) → DONE in wetSpring Exp273–286 + neuralSpring nS-601–605 (359/359 checks). Oclacitinib JAK selectivity, IL-31 pruritus, lokivetmab PK, three-compartment tissue, cross-species

Tier 3 — Longer horizon, connect to broader themes

17. Bazavov et al. (2025) — Phys Rev D — Muon g-2 HVP → precision lattice computation
18. Fernandez et al. (2020) — PNAS — Cell shape regulation → bistable dynamical systems
19. Dolson (2023) — J R Soc Interface — Ecology-evolution continuum → origin-of-life context
20. Liu working manuscript — Fungi-bacteria coevolution → cophylogenetic methods for wetSpring
21. Filonov & Kachkovskiy (2018) — Acta Math — 2D band edge structure → electronic/phononic transport, deep eigenvalue mathematics
22. Mateos, Anderson et al. (2023) — Science Advances — Sulfur-cycling enzyme evolution across geological time → deep-time validation of constrained evolution, bioinformatics pipeline
23. Boden, Anderson et al. (2024) — Nature Communications — Phosphorus-cycling enzyme timing → tree reconciliation methods, geological time enzyme dating

Related

• Lab — reproduction infrastructure and validation summaries
• Chapter 14: Biological Validation — empirical support for constrained evolution
• For Faculty and PIs — audience guide for faculty engagement